OpenAI releases GPT-5.6 models after government review, unveils ChatGPT Work to compete in AI agent race

OpenAI's GPT-5.6 Is Dropping on Thursday: What's Different About Sol, Terra and Luna

OpenAI is getting ready to release its next generation of AI models. The GPT-5.6 series, which includes a trio of new models, will be released on Thursday, CEO Sam Altman confirmed in a post on X Wednesday. The flagship model is called GPT-5.6 Sol. It's meant to handle deep reasoning tasks and manage agents. Terra is the middle, Goldilocks option in terms of capability and cost, while Luna is the quickest and cheapest to run. They are a "meaningful step up in cybersecurity capabilities" but also have advanced safety guardrails, the company wrote when outlining the model family in June. OpenAI said Sol makes fewer factual errors, or hallucinations, compared to previous OpenAI models. We don't yet know the specific availability of these models, other than all three will be released on Thursday. CNET reached out and did not immediately receive a response from OpenAI. While we've known about GPT-5.6 since its announcement on June 26, OpenAI has purposefully slow-rolled its release. Like Anthropic's Claude Mythos 5, the new GPT models are purportedly the most advanced AI available, and because of that, AI companies are being more hesitant with making them public. And for the first time, the US government is holding companies back with new requests for government review of AI models before they can be widely released. On June 2, President Donald Trump issued an executive order requiring the Defense Department to design a system that has AI developers voluntarily sharing access to new frontier models with the government. Government officials have 30 days to review the tech and raise red flags. (This is a different legal standard than the one under which the government killed and then resurrected Claude Fable 5.) OpenAI said that it had voluntarily given the government and some "trusted partners" early access. It's not clear who these partners are and what their expertise is, and OpenAI declined in June to comment on the specifics, referring CNET to the announcement. A White House official told CNET that the US government did not give OpenAI the green light to release GPT-5.6, nor was that required before a public release under the executive order. They emphasized that all engagement is voluntarily. This is a big change in the government's stance on AI. The Trump administration had long been adamant that regulation would slow innovation and cause the US to "lose" the global AI race to China. Now, with government oversight on AI ramping up, it changes the pace of model release and the US's technological reputation on the world stage.

OpenAI rolls out GPT-5.6 after government green light -- and announces 'ChatGPT Work'

About two weeks after OpenAI's GPT-5.6 was caught up in regulatory drama -- rolled out only to government-approved organizations during a "limited preview" period -- the company has received the Trump administration's green light for a public rollout of the model. To celebrate, OpenAI also unveiled a new AI agent on the same day: ChatGPT Work. It's billed as a combination of ChatGPT and Codex, allowing the everyday non-technical user to take advantage of Codex's capabilities for non-coding tasks, and it's powered by the GPT-5.6 model suite (Sol, Terra, and Luna). "It can gather context from the apps, files, and workflows you choose and create finished materials such as documents, spreadsheets, presentations, and web apps," OpenAI wrote in a blog post, adding that a "unified plugins directory" allows ChatGPT to connect to tools like Slack, Gmail, Google Drive, calendars, and CRMs. Mac and Windows users worldwide, including free ChatGPT users, should have immediate access to ChatGPT Work and GPT-5.6 via the ChatGPT desktop app. On mobile and the web, Pro, Enterprise, and Edu users will first get access, while Plus and Business users will receive access "over the next few days," OpenAI wrote, adding that the "rollout is starting globally and will continue gradually toward full availability over the next 24 hours." Companies like OpenAI and Anthropic, as well as tech giants like Google (and even, recently, Apple), have been vying to clear new ground in the race to make AI agents actually useful for the average person, especially in the wake of the viral open-source AI agent OpenClaw. They've had varying results, and for now, the theoretical right-hand AI agent for the everyday consumer remains out of reach. OpenAI is hoping that its new product, which is a direct competitor to Anthropic's Claude Cowork (combining its own Claude and Claude Code), will push it ahead in the race. OpenAI is especially banking on Sol, the most powerful of the GPT-5.6 model suite, to set "a new standard for intelligence and efficiency," particularly when it comes to coding, cybersecurity, and science, as well as computer use capabilities. The company is also marketing the model as a lower-cost alternative to competitors' most powerful models, amid complaints of an industry-wide money squeeze and AI lab costs being passed onto customers.

OpenAI to release delayed models Thursday amidst a sea of regulatory confusion

After the administration initially requested restriction of access to the models, a US government statement now says that no permission for their release was ever required. As enterprises struggle to manage their AI strategies, the US AI regulatory environment is sending a wide range of contradictory signals. OpenAI's Wednesday announcement that it will now release GPT-5.6 Sol, along with Terra and Luna, on Thursday highlights the confusion. Initially, the US government said that it was asking OpenAI to limit access to its top models, including the three releasing Thursday, to a short list of companies. OpenAI seemingly agreed and held back their general availability. But on Wednesday, OpenAI reversed its position, with a statement on X saying simply: "GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We're expanding preview access globally now." No details were released about the extent of the expansion.

OpenAI's newest AI model is 54% more token efficient on agentic coding, Altman tells CNBC

OpenAI CEO Sam Altman told CNBC on Thursday that GPT-5.6 Sol, the company's latest artificial intelligence model, is 54% more token efficient on agentic coding tasks, and that it's "as good or better" than competing models on the market. "Every enterprise now is thinking about spend and the value they're getting in exchange for AI, and this is what we really want to do," Altman said. OpenAI on Thursday is rolling out its latest series of models, GPT-5.6 Sol, Terra and Luna, which it announced last month. The company limited the initial launch to a "small group of trusted partners" at the request of the U.S. government. Altman said the company worked with Commerce Secretary Howard Lutnick, Treasury Secretary Scott Bessent and U.S. National Cyber Director Sean Cairncross on the approval process. He described OpenAI's work with the government as a "collaborative back and forth," where the government would carry out tests and raise problems for the company to address. "If you want broad access, which we do, and you have powerful models, you really want to be able to be confident in your safety claims, because otherwise the world is going to get uncomfortable very fast," Altman said.

OpenAI gets permission to roll out GPT-5.6 to the public on July 9 - Engadget

OpenAI will publicly launch all three of GPT-5.6's variants -- Sol, Luna and Terra -- this Thursday, July 9, the company has announced on X. "We're expanding preview access globally now," it added. The company initially made the new model series available to a "small group of trusted partners" in late June, due to a request from the Trump administration. If you'll recall, President Trump had signed an AI cybersecurity order in early June, which asks companies to voluntarily present their most powerful models for government review 30 days before releasing them to the public. For GPT-5.6's initial release, it was only rolled out to select government-approved entities. "We don't believe this kind of government access process should become the long-term default," OpenAI said back then, but it added that it was complying because it was the best way to ensure that it can release its latest model series to the public soon. Sure enough, it will not take OpenAI 30 days before making GPT-5.6 more broadly available. According to Axios, the Trump administration gave OpenAI permission for a wider release after putting the model through additional testing and conducting more meetings with the company. The Department of Commerce's Center for AI Standards and Innovation conducted those additional tests, with the company sending technical experts to DC so they can immediately address potential questions and concerns. GPT-5.6 comes in three variants, with Sol being the company's strongest model yet. Terra is for everyday use and promises a similar performance to GPT‑5.5 despite being twice as cheap, while Luna is the company's lowest cost model. The Sol variant costs $5 per million input tokens and $30 per million output, whereas Terra costs $2.50 per million input tokens and $15 for output and Luna costs $1 per million input and $6 for output. Anthropic also had to block all access to its latest Mythos cybersecurity and Fable models to ensure compliance when the government ordered it to prevent all foreign nationals from being able to use them. The company has since received the government's permission to redeploy the more exclusive Mythos 5 model and, eventually Fable 5, its counterpart that brings many of its capabilities to the public.

OpenAI Releases GPT-5.6 Sol, Its Most Powerful AI Model Yet

The new release was delayed after the U.S. government restricted the latest artificial intelligence models over cybersecurity concerns. OpenAI has publicly released its flagship artificial intelligence technology, GPT-5.6 Sol, its most powerful model yet, alongside a new tool to assist with everyday office work. The new release, on Thursday, escalates the battle for A.I. dominance with the rival start-up Anthropic as the technology grows increasingly sophisticated. Along with its new model, OpenAI unveiled a new tool, ChatGPT Work, which is powered by Sol. Much like Anthropic's Claude Cowork, ChatGPT Work is an "A.I. agent" that can operate software apps and websites on behalf of users, including spreadsheets, online calendars and email services. When running standard benchmark tests, OpenAI's new flagship model roughly matches the performance of Anthropic's leading model, Fable 5, according to Vals AI, a company that tracks the performance of the latest A.I. technologies. "In real-world situations involving financial and legal tasks, it is state of the art," said Rayan Krishnan, the chief executive of Vals AI. The release of the more powerful model followed U.S. government efforts to restrict both OpenAI's and Anthropic's new A.I. models over cybersecurity concerns. While President Trump had long taken a hands-off approach in regulating the A.I. industry, he signed an executive order last month that asked tech companies to voluntarily give the government oversight of new A.I. models before releasing them to the public. Last week, OpenAI said it would share its new technology with only a small group of companies approved by the Trump administration. But after lengthy discussions, the administration allowed both OpenAI and Anthropic to release their new technologies publicly. Although the administration initially treated the technologies as threats to national security, many independent security experts said the dangers were overstated. OpenAI is taking a more open approach to cybersecurity than Anthropic, which has tightly restricted the use of Fable to prevent hackers from using its technology. OpenAI's new model includes fewer guardrails, so that individuals and businesses can use the technology to defend themselves. Anthropic released a new A.I. system called Claude Mythos in April, but said it couldn't share the technology with the public. Some executives and government officials worried that the technology would usher in a new era of cybersecurity threats. It released Fable 5, a more limited version of Mythos, last month. Days later, the Trump administration ordered Anthropic to suspend access for all foreign nationals, citing national security concerns. To comply with the order, Anthropic took the technology offline entirely. The company has added guardrails to Fable designed to block responses related to cybersecurity, biology and other vulnerable areas. When Fable blocks responses, it sends them to a less powerful technology, Claude Opus 4.8. Technologies like Fable and Sol are particularly good at identifying vulnerabilities in the software that underpins the internet. This means that malicious hackers can use the technology to attack computer networks. But it also means that businesses and governments can use it to defend against these attacks. Fable's guardrails may prevent hackers from using the technology for offensive purposes. But the same guardrails may also bar businesses and governments from using the system for defense. Like Anthropic's Fable, OpenAI's GPT-5.6 Sol is expensive compared with less powerful models. Fable is roughly twice as expensive as Anthropic's next most powerful model, and OpenAI's model is more expensive than Fable in some cases, Mr. Krishnan said. On Thursday, Meta also released a new A.I. technology. It was not as powerful as GPT-5.6 Sol or Fable, and is also considerably less expensive. (The New York Times has sued OpenAI and Microsoft, claiming copyright infringement of news content related to A.I. systems. The two companies have denied those claims.)

OpenAI releases GPT-5.6 after US government approval

The public launch ends weeks of government-gated access and sets a first: an American frontier model cleared for release through a federal review OpenAI has made GPT-5.6 publicly available after receiving approval from the Trump administration. The release ends weeks in which the model was restricted to a small group of government-vetted partners, CNBC reports. The model family comes in three tiers. Sol is the flagship, Terra a lower-cost enterprise option, and Luna the fastest and cheapest. OpenAI says Sol delivers 54% greater token efficiency on agentic coding tasks, putting it level with or ahead of rivals. Its strengths in coding, biology, and cybersecurity were exactly what drew federal scrutiny in the first place. The review ran through the Commerce Department's Center for AI Standards and Innovation. Commerce Secretary Howard Lutnick, Treasury Secretary Scott Bessent, and National Cyber Director Sean Cairncross were involved, in what OpenAI described as a "collaborative back and forth". Sam Altman said OpenAI made "many changes" during the talks, and sent technical staff to Washington to answer the agency's questions. This was the first US frontier model launched under a government-managed access list. How a model ended up needing a permission slip The gating was not routine. Washington had asked OpenAI to stagger the release because of the model's advanced capabilities, echoing a wider push to vet powerful systems before they ship. A recent Trump executive order asks AI firms to voluntarily give the government up to 30 days of pre-release access to models with advanced cyber abilities. GPT-5.6 became the first big test of that framework. The precedent was set next door, at Anthropic. Its powerful Mythos and Fable 5 models were pulled offline over cyber-capability fears, and the company faced export controls that were only recently lifted. A safety win, or a worrying precedent Supporters see a responsible system working, since models that can find software flaws or aid bioweapon design warrant a hard look before public release. A collaborative review, on this reading, beats either reckless launches or outright bans. Critics see something more uncomfortable, namely the executive branch deciding when a private company may ship a product. Altman himself reportedly told staff the government-curated access list was not a sustainable long-term approach. The unease is sharpened by the vagueness of the rules. Microsoft's Brad Smith has argued the US now regulates AI without clear, transparent rules, leaving companies to negotiate case by case. For now, GPT-5.6 is out, and the framework held together on its first real run. Whether case-by-case clearance becomes the norm, or a friction founders eventually route around, is the question this launch leaves open.

White House Denies Giving OpenAI 'Green Light' to Publicly Release Its Latest Model

There seems to be some confusion around who has the authority to approve the release of new AI models in the United States. OpenAI announced in an X post on Tuesday night that its latest model, GPT-5.6 -- which comes in three tiers: Sol, Terra, and Luna -- will be publicly available starting tomorrow, July 09. The model was first introduced last month, with OpenAI announcing in a blog post that it would initially be rolled out solely to a small group of government-approved partners. Previously, company CEO Sam Altman had reportedly told employees in an internal memo that the government would be granting access to the model on a "customer by customer" basis. Axios reported this morning, citing an anonymous source, that the decision to make Sol, Terra, and Luna publicly available followed a "green light" from the Trump administration. OpenAI has been collaborating with the Department of Commerce's Center for AI Standards and Innovation to test the new model for potential national security risks, and the company has sent technical experts to Washington DC, where they'll remain to continue working with federal officials, the outlet said. The implication from the Axios report is that OpenAI required some kind of official permission before it could proceed with the public release of GPT-5.6. If that's in fact the case, then it seems to contradict the Trump administration's own (at least public) directive on how AI models can be deployed in the U.S. In an executive order published June 02, the president proposed a safety testing framework through which AI companies would voluntarily grant federal officials access to new models for 30 days before public release. The framework, according to the order, would under no circumstance include "a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models, including frontier models." Following that party line, a White House spokesperson later told Axios that the administration had not given OpenAI any kind of official approval to make GPT-5.6 publicly available, since -- as Trump's June 02 executive order makes plain -- no such green light is required. The decision-making processes around when, how, and to whom new models are released "rest entirely with the companies," the spokesperson told Axios. The federal government, however, has plenty of levers it can theoretically pull if and when it wants to delay or halt the release of a particular model. This was, of course, most brazenly illustrated last month, when the Department of Commerce ordered Anthropic to block access for all "foreign persons" to the company's newest models, Fable 5 and Mythos 5, ostensibly due to national security concerns. The order effectively forced Anthropic to take both models offline for all customers and was met with widespread confusion and frustration from both cybersecurity and legal experts. When a modified version of Fable 5 -- one with even more restrictive cybersecurity guardrails than the original -- was released last week, Anthropic said its own internal testing had shown that the cybersecurity capabilities that had allegedly prompted the Trump administration's export ban could also be demonstrated by a host of other, already generally available models, including earlier versions of Claude and OpenAI's GPT-5.5. The White House and OpenAI did not immediately respond to requests for comment.

OpenAI shares update on GPT-5.6 availability after holding back release

OpenAI unveiled GPT-5.6 at the end of June, but the public release was held back while the U.S. government reviewed the new AI models. Now the company has shared an update on when customers will be able to use the upgraded model: this week. GPT-5.6 is coming on Thursday To recap, OpenAI announced three versions of GPT-5.6: * Sol, OpenAI's flagship model * Terra, their balanced model for everyday work * Luna, their fast and affordable model The Sol, Terra, and Luna naming scheme is also new with GPT-5.6. "In this new naming system introduced with GPT‑5.6, the number identifies a model's generation, while Sol, Terra, and Luna identify durable capability tiers that can advance on their own cadence," OpenAI said in June. "Together, the family gives people and developers clearer choices across intelligence, speed, and cost." OpenAI also said it would only share access to GPT-5.6 models with a small group of trusted partners at the U.S. government's request. This marked a shift from OpenAI's previous announcement and release behavior. Now OpenAI has confirmed that GPT-5.6 Sol, Terra, and Luna will become available Thursday, July 9. OpenAI also said it was "expanding preview access globally now" in a social media post on Tuesday night. It's unclear if tomorrow's GPT-5.6 release will be limited to users in the United States or globally available. OpenAI's Sam Altman said last month that the company was working toward a global release.

OpenAI wins US clearance for a broad GPT-5.6 rollout after weeks of government testing

The Commerce Department has signed off on a wide release, ending a restricted preview that had held the model to about 20 vetted partners. OpenAI has been cleared to release its most advanced model widely, after the US government signed off on a broader rollout of GPT-5.6 that had been held back for weeks under Washington's new oversight regime for frontier AI. Until now the model had been available only through a restricted preview to about 20 partners whose names were individually approved by the US government. That arrangement, the first of its kind for an American frontier model, is what the wider release now supersedes. The sign-off followed additional testing by the Commerce Department's Center for AI Standards and Innovation, the body set up to vet advanced systems. OpenAI sent technical experts to Washington to answer the agency's questions, according to Axios. GPT-5.6 is a three-tier family rather than a single model. Sol is the flagship, Terra a lower-cost mid-tier option, and Luna the fastest and cheapest of the three. OpenAI has described Sol as strong at coding, biology and cybersecurity, and paired it with a "max reasoning effort" mode that gives the model more time to work through hard problems. Those same capabilities, particularly in biology and cyber, are part of why the government wanted a closer look before a wide release. The tiering is a commercial choice as much as a technical one. Terra is aimed at everyday enterprise workloads where cost matters more than raw capability, while Luna is built for high-volume tasks that need speed above all, a split that lets OpenAI charge very different prices across the same family. The preview it now supersedes was unusually tight. For weeks GPT-5.6 was available only to a short list of organisations whose identities OpenAI had shared with the government, the first time an American lab had gated a frontier model behind a state-approved roster. The review sits inside a framework the Trump administration established on 2 June, which introduced a voluntary pre-release check for the most capable models. The GPT-5.6 case went further than that, moving from voluntary review to a government-managed access list, a step OpenAI had agreed to only after being asked to slow the launch. OpenAI has made clear it is uneasy with the precedent. The company said it does not believe this kind of government access process should become the long-term default, while agreeing to take part this time. The discomfort is not hard to understand. A government that can gate a launch can also stop one, a power the administration has already used elsewhere in the sector by ordering Anthropic to shut down two models. For OpenAI, the commercial stakes of the delay were real. Every week that GPT-5.6 stayed inside a 20-partner preview was a week rivals could court the enterprise customers it wanted to reach with the new tiers. The company now expects to widen access to GPT-5.6 within days, building on the base it laid with GPT-5.5 earlier in the year. It has said all three tiers will become generally available in the coming weeks, though it has not fixed a public date. OpenAI is not the only lab inside the new regime. The same framework covers its rivals, which means the way this rollout has gone is likely to shape how the next frontier model from any US company reaches the public. What the episode establishes is less a product timeline than a template. For the first time, a leading US lab has released a frontier model on the government's schedule rather than its own, and both sides now have to decide whether that was a one-off or the shape of things to come.

GPT-5.6 Sol, Terra, and Luna are here. See which one's best for you.

Before building anything, let's first figure out which model to use. Credit: OpenAI/YouTube These days, new versions of AI chatbots don't just launch; they're unshackled and released to the public following government scrutiny. OpenAI's new GPT-5.6 models were - like Anthropic's Claude Mythos and Fable - apparently too powerful to just launch; but now, after some tinkering, they're available to you, dear customer. In practice, it simply means that the new GPT-5.6 models are very powerful and smarter than before. In its introductory post, OpenAI shared a bunch of graphs showing just how much better GPT-5.6 is than the competition, whilst using fewer tokens and generally costing less. OK, great. But GPT-5.6 is not just one model; it comes in three distinct flavors: Sol, Terra, and Luna. So what do different kinds of users get, what should they pay for, and which models should they (mostly) use? Let's dive in. Free users get (almost) nothing Sorry; if you're not a paying customer, you'll have to make do with OpenAI's previous flagship model, GPT-5.5. Any sort of access to GPT-5.6 models requires a subscription of some sort. Fortunately, GPT-5.5 is still quite capable at most tasks, but if you want the best of the best, you'll have to cough up the dough. There's an exception to this: Free and Go users can access GPT-5.6 through ChatGPT Work. More on that below. If you're a Plus or Business user, you can only get Sol (the most powerful model) at medium and higher effort settings. There's another, higher level of performance called Sol Pro, but that's only available for Pro and Enterprise users. In terms of availability per one million tokens, the prices are: $5 input and $30 output for Sol. $2.5 input and $15 output for Terra, and $1 input and $6 output for Luna. Sol, Terra, or Luna? Why are there three models in the first place? Well, OpenAI always had a multi-tier model; for example, previously users were able to choose a "mini" version of the main model to get results done cheaper. Now, the model has been split into three tiers. If you're a paying customer, you're free to use all three. But you know how it is in the world of LLMs: If you pick the smartest one, your usage limits will get hit faster (yes, there are always usage limits, even if you throw a ton of money at OpenAI). In the simplest sense, GPT-5.6 Sol is the smartest model, Terra is in between (with roughly GPT-5.5 level of performance), and Luna is the cheapest, fastest, but also least capable of the bunch. The breakdown is as follows: Terra is a "balanced" model for everyday work. That's the one you should be asking most of your questions. Don't underestimate it, though, as OpenAI claims it performs better than Anthropic's Fable 5 in some cases. Luna is cost-efficient, and should be used for easy, non crucial everyday tasks; think recipes and movie recommendations. Again, OpenAI says it outperforms Anthropic's Opus 4.8 in some cases, so it's not a slouch, either. Sol should be reserved for coding, deep research, planning, and cybersecurity: The most demanding tasks. Of course it comes at a (literal) cost: While OpenAI claims it spends less tokens than Anthropic's Fable 5, Sol will still hit usage limits a lot faster than the other variants. Fun fact: If you ask GPT-5.5 about any of this, right now, it'll give you completely wrong answers. Hopefully OpenAI will fix this soon. Wait, what's this ChatGPT Work thing, then? Oh yeah, OpenAI also launched ChatGPT Work, which is a new agent in ChatGPT that can access and take actions on your apps and files, and work in the background until a task is finished. It's powered by Codex (OpenAI's software engineering agent) and GPT-5.6. Think about it as your buddy that will go through your emails and files, browse the web, fetch the relevant data, and create that presentation your boss wants before the day is done. ChatGPT Work is rolling out to Pro, Enterprise and Edu users first on web and mobile; this will be expanded to Plus and Business users "over the next few days." On the desktop, Work is available for everyone, including Free users. Oh, and one more thing: The fact that ChatGPT Work has a built-in browser also means that OpenAI is sunsetting its standalone web browser, Atlas. Sorry. How about GPT Live? GPT Live is a new version of ChatGPT Voice and it will show up when you start talking to ChatGPT. We've covered this in more depth here, but the bottom line is that GPT Live can listen and speak at the same time, allowing it to keep up a more realistic conversation. Wrapping it all up The new GPT-5.6 model is smart. It comes in three flavors: Luna, Terra, and Sol, with Sol being the most capable variant, Luna the most affordable one, and Terra somewhere in the middle. You can currently only get them on paid tiers, unless you're using ChatGPT Work on desktop. And ChatGPT Voice has also gotten smarter with GPT-Live underneath, a model that can listen and speak at the same time.

OpenAI releases latest ChatGPT model after delay over White House cybersecurity concerns

Staggered release of ChatGPT 5.6 follows similar restrictions on rival firm Anthropic's latest AI models OpenAI released its latest advanced AI model, called ChatGPT 5.6, on Thursday after earlier delaying the public rollout over US government concerns about cybersecurity. The Trump administration had requested last month that OpenAI limit the release to a small group of government-approved users. OpenAI complied with the White House's request last month. The company stated in a blogpost that it had briefed government officials on ChatGPT 5.6's capabilities and restricted the model to trusted partners at their behest. The product's wider release came after additional testing by the government's Center for AI Standards and Innovation agency, according to Axios. The staggered release of ChatGPT 5.6 follows similar restrictions on rival firm Anthropic's latest AI models, also put in place by the Trump administration last month, which resulted in a temporary export ban on the products. Both episodes have brought attention to the cybersecurity risks associated with AI, tensions with US allies over sharing breakthroughs and questions over the US government's slapdash process for regulating the technology. OpenAI's ChatGPT 5.6, which includes a flagship product called Sol, claims to be its safest and most capable model yet. It is the company's answer to Anthropic's Claude Fable and Mythos models, which rolled out to the public earlier this month after government restrictions. The Trump administration has largely spurned domestic calls for reining in the AI industry and encouraged companies to advance their models as quickly as possible, often citing an AI arms race with China and the need for US dominance over technology. More recently, the threat of destabilizing, AI-driven cyber-attacks has caused the White House to increase its intervention, with Donald Trump issuing an executive order last month calling for a voluntary review process for AI firms to submit their latest models to the government for evaluation. The US restrictions on who can access the most advanced AI models has caused concern within industries that rely on the models to do business, adversarial governments that face potential cybersecurity threats and allies that felt snubbed by the US' export ban. OpenAI initially only gave access to its latest model to US entities. Anthropic and OpenAI are jostling for the lead in the AI race ahead of their initial public offerings eyed for later this year. Anthropic announced a new round of funding in May that made it the world's most valuable AI company at a valuation of $965bn. OpenAI is reportedly seeking a valuation of $1tn for its own IPO, which the New York Times reported could be delayed until next year amid turmoil in the global tech market.

OpenAI releases GPT-5.6 and ChatGPT Work tool

Why it matters: The leading players in AI are all speeding forward at breakneck speed, but the increasing power of the latest models has the Trump administration rethinking its previous hands-off approach. Driving the news: OpenAI is releasing three flavors of GPT-5.6: * Sol is the most powerful version, while Luna is designed for speed. Terra, meanwhile, aims to strike a balance between the two for everyday work. * A new "ultra" mode within Sol allows the system to work even harder on tasks and to delegate work to various submodels. * CEO Sam Altman told CNBC that Sol is 54% more token efficient on agentic coding tasks, an important stat as "every enterprise now is thinking about spend and the value they're getting in exchange for AI," he said. Zoom in: Alongside the new models, the company also launched ChatGPT Work on Thursday. * The new agent, powered by 5.6, can gather context across connected apps and files to create documents, spreadsheets, presentations and other work. * It can also work across web, phones and computers, first rolling out to the Mac and Windows apps for all tiers of users, with web to follow. The big picture: The model release comes after the administration asked OpenAI to delay its rollout. * Altman said the company made "many changes" after a "collaborative back and forth" with the administration. * He called the government's technical capabilities "impressive." Friction point: OpenAI has been battling Anthropic for the crown of having the most powerful model. * At the same time, Meta, SpaceXAI and others have been aggressively trying to catch up with the leaders. Both companies released updated models this week. Between the lines: GPT-5.6 is already getting rave reviews from many of those who have been testing it. Some say that they see Anthropic's Fable as having greater raw intelligence, but GPT-5.6 is viewed as being a more reliable model for regular tasks. What they're saying: MagicPath AI CEO Pietro Schirano wrote on X: "I've been testing it for months and, without exaggeration, it's the best model I've ever used. Fast, smart, genuinely creative." * Theo Browne, CEO of chatbot platform T3 Chat, wrote on X: "gpt-5.6-sol is world-leading in computer use. It made me use it 100x more. When we lost access to 5.6, I quickly started to go insane without it." The other side: Some who have tested both Fable and GPT-5.6 expressed a preference for Anthropic's model.

If you've grown tired of babysitting ChatGPT, the new GPT-5.6 models might be the fix

OpenAI seems to have a new AI model waiting in the wings every few months, and today is no different. The company has officially unveiled the GPT-5.6 family, bringing three new models to ChatGPT, Codex, and its API. The big star of the show is GPT-5.6 Sol, but it's joined by Terra and Luna, which are designed to deliver strong performance at a lower cost. The days of endless follow-up prompts may be numbered If you've ever felt like ChatGPT needs too many follow-up prompts to finish a job, OpenAI thinks this update can help. Instead of simply answering one question at a time, GPT-5.6 is designed to handle bigger, multi-step tasks with less hand-holding. For example, imagine you're planning a weekend trip. Rather than asking ChatGPT where to go, then where to stay, then what to do, and finally asking it to put everything together, GPT-5.6 is supposed to handle much more of that work on its own. The same idea applies to coding projects, research, spreadsheets, or even comparing dozens of products before recommending the best option. The flagship GPT-5.6 Sol model is built for those heavier workloads, and OpenAI says it delivers better performance while using fewer tokens than previous models. That means it can accomplish more work without driving up computing costs, which is good news for developers and businesses that rely on OpenAI's models every day. One of the biggest additions is a new Ultra mode. Instead of relying on a single AI process, Ultra can split a complicated task across multiple AI agents working in parallel. You can think of it like assigning research, writing, editing, and fact-checking to four people instead of asking one person to juggle everything. According to OpenAI, this helps solve difficult problems faster while improving the final results. There's a GPT-5.6 for every kind of job For users who don't need that much power, OpenAI is also introducing GPT-5.6 Terra and GPT-5.6 Luna. These models are designed to be more affordable while still handling common AI tasks well. The company says both outperform competing models in their respective categories while costing significantly less to run, making them attractive options for developers building AI-powered apps. The models are also becoming more independent. Instead of stopping after every instruction, GPT-5.6 can write small programs, use tools, process information, check its own progress, and decide what to do next with fewer prompts from the user. That could make tasks like debugging code, organizing research, or pulling together reports feel much smoother. If accuracy matters more than speed, OpenAI is also adding a Max mode. It gives GPT-5.6 extra time to think through difficult questions, test different approaches, and double-check its work before responding. Ultra goes even further by using multiple AI agents simultaneously, trading extra computational power for better results on more demanding jobs. OpenAI also says it put GPT-5.6 through its most extensive safety testing yet, combining human red-team exercises with automated evaluations to make the models more resistant to misuse without impeding legitimate use. Recommended Videos The GPT-5.6 family is rolling out starting today across ChatGPT, Codex, and the OpenAI API, with global availability expected to expand over the next 24 hours. For most ChatGPT users, the benchmark numbers probably won't mean much. What will matter is whether GPT-5.6 can actually save you time by needing fewer prompts, handling larger tasks on its own, and delivering answers that require less back-and-forth. If OpenAI's claims hold up, that could be the biggest upgrade of all.

OpenAI launches GPT-5.6 Sol after U.S. government approval

OpenAI began rolling out GPT-5.6 Sol to the general public on Thursday after the Trump administration approved a broad release, ending weeks of restricted access to a small group of government-vetted partners. Speaking to CNBC, CEO Sam Altman said Sol achieves 54% greater token efficiency on agentic coding tasks, putting it at parity with or ahead of rival models. "Every enterprise now is thinking about spend and the value they're getting in exchange for AI, and this is what we really want to do," Altman said. GPT-5.6 is a family of three models. Sol is the flagship tier, Terra is a lower-cost option aimed at enterprise workloads, and Luna is the fastest and least expensive of the three, designed for high-volume tasks. Altman said the approval process involved Commerce Secretary Howard Lutnick, Treasury Secretary Scott Bessent, and U.S. National Cyber Director Sean Cairncross. He characterized the dynamic as a "collaborative back and forth" in which federal reviewers flagged concerns for OpenAI to resolve. "If you want broad access, which we do, and you have powerful models, you really want to be able to be confident in your safety claims, because otherwise the world is going to get uncomfortable very fast," he said. The GPT-5.6 rollout had been limited to roughly 20 partners after the Trump administration asked OpenAI to stagger the release while federal agencies evaluated the model. The Commerce Department's Center for AI Standards and Innovation carried out the evaluation, and OpenAI sent technical staff to Washington to respond to the agency's questions, according to prior reporting. Sol's capabilities in coding, biology, and cybersecurity were among the factors that drew federal scrutiny. The restricted rollout went beyond a voluntary pre-release review framework President Trump signed on June 2, which called on companies to allow agencies to check powerful AI systems before launch. Where that structure relied on voluntary participation, the GPT-5.6 situation resulted in a government-curated access list -- an arrangement Altman told employees OpenAI did not view as a sustainable long-term approach. The situation drew comparisons to Anthropic's recent experience, in which a Commerce Department export control order forced the company to pull both its Fable 5 and Mythos 5 models from general availability. Fable 5 has since been reinstated, but Mythos 5 remains restricted to a narrow set of American organizations, according to prior reporting. Altman confirmed that discussions with the Trump administration over a potential government equity position remain ongoing, pushing back on a Financial Times report pegging the proposed stake at 5% and calling the account riddled with "a lot of inaccuracies."

OpenAI Releases GPT-5.6 Sol: Here's How It Stacks Up Against Other AI Models

OpenAI's GPT-5.6 Sol is now public. The company released its new flagship model to general users today, launching alongside two smaller siblings, Terra and Luna, after the U.S. Department of Commerce kept a preview restricted to about 20 trusted partners for two weeks. The naming strategy is new for OpenAI. This is the first time the company gives names to its models instead of simply using numbers. Sol, Terra, and Luna mark capability tiers that can move on their own cadence. Sol is the flagship, Terra the everyday model OpenAI says matches GPT-5.5 at half the price, and Luna the cheaper option. Pricing runs $5 and $30 per million input and output tokens for Sol, dropping to $1 and $6 for Luna. (Tokens, for those not in the know, are the smallest unit of information a model can handle. And companies typically price their models on a per token basis for API services) Two new knobs ship with it: a max reasoning effort that lets Sol think longer, and an ultra mode that farms work out to subagents. For context, Anthropic charges $10/$50 for Claude Fable 5, Google charges $2/$12 for Gemini 3.1 Pro, xAI charges $15/$75 for Grok 4.5. On the Chinese side: DeepSeek charges $1.74/$3.48 for V4 Pro, and Xiaomi charges just $1/$5 for MiMo v2.5 Pro -- placing Sol between the premium U.S. frontier models and China's low-cost challengers. What the benchmarks show On Terminal-Bench 2.1 -- a test of command-line workflows that reward planning, tool use, and iteration, scored as the share of tasks a model completes -- Sol in its ultra configuration hit 91.9%, with standard Sol at 88.8%. That puts both ahead of Anthropic's Claude Mythos 5 at 88.0%, Claude Fable 5 at 84.3%, and Claude Opus 4.8 at 78.9%. Google's Gemini 3.1 Pro Preview trailed the chart at 70.7%. OpenAI leaned hardest on cyber. On ExploitBench, which measures how well a model finds and weaponizes software vulnerabilities, Sol matched the restricted Mythos Preview while spending roughly a third of the tokens. OpenAI says Sol still doesn't cross the "Cyber Critical" line in its own risk framework. The testers already have opinions Early access was loud. Theo, a well-known developer, AI youtuber and CEO of the AI platform T3 Chat, called Sol "world leading in computer use" and said it fixed the complaints he had with GPT-5.5. Dan Shipper, whose team at Every tested it for a month, offered the cleaner line: "GPT-5.6 is like a Porsche, Fable is like a warp drive." His read is that Sol is the daily driver and Fable the thing you reach for to cross the galaxy. Researcher Daichi Konno, who got in early, said Sol clearly beats GPT-5.5 and lands near Fable 5, with Anthropic still ahead in writing tasks. His sharper note: Sol's safeguards didn't trip on life-science questions, which he thinks could make it a default for biology work. Leaks already point past it. One roadmap-tracking account, "Synthwave" on X, claims GPT-5.6 is the last of the 5.x line, with GPT-6 built on a larger base arriving within about a month. The same thread pegs Anthropic's Fable 5.1 as close and DeepSeek's V4 general release as imminent. The timing is pointed. Sol arrives the same week Anthropic's Fable 5 drops out of subscription plans -- after it returned globally on July 1, the model moved to usage credits only once its 50% weekly allowance expired on July 7. It's crowded at the top. SpaceXAI shipped Grok 4.5 yesterday at a fraction of the price, which Elon Musk called "roughly comparable to Opus 4.7, but much faster." Meta released Muse Spark 1.1 this morning, its first paid model. Neither leads the pack, but both are frontier-class. That leaves Google as the one big U.S. lab that hasn't refreshed its flagship in this run. Its top model, Gemini 3, has been out since November 2025 -- the oldest frontier release still standing while OpenAI, Anthropic, xAI, and Meta all moved in the same seven days.

OpenAI to launch new model after US freeze

San Francisco (United States) (AFP) - ChatGPT maker OpenAI said its latest powerful artificial intelligence model series will be released to the public on Thursday, as the US government reportedly approved a broader launch. The company's new GPT-5.6 offerings and other cutting-edge AI models, including Anthropic's Mythos series, have drawn concern over their supposedly unprecedented ability to identify weaknesses in code that hackers can exploit. That has raised national security fears, and OpenAI said in late June it had shared preview access to GPT-5.6 with a limited group of trusted US-based partners at Washington's request. Large language models are the technology that underpins chatbots and many other AI tools, with their capacity to crunch through colossal amounts of digital data. The GPT-5.6 series has three tiers: Sol, the company's new flagship model; Terra, a mid-range version for everyday work; and Luna, a fast, low-cost option. "GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We're expanding preview access globally now," OpenAI said in an X post Tuesday, without giving further details. US news outlet Axios reported, citing a source familiar with the situation, that the Trump administration had given the company the green light for a broad launch of GPT-5.6, following technical testing and meetings between the company and government officials. AFP has contacted OpenAI, the White House and the US Department of Commerce for comment on the Axios report. It follows a similar story at OpenAI's archrival Anthropic, the startup behind the Claude chatbot. Last week, Anthropic said it would begin restoring access to its most powerful AI models, Fable 5 and Mythos 5, after Washington lifted a restriction on where they could be released. Before Mythos's arrival, President Donald Trump's administration wanted fewer rules on AI companies, not more, hoping that would help the US beat China in the AI race. The government is now drawing up criteria for which AI models would fall under new security restrictions, in accordance with an executive order from the White House. OpenAI said in June that "we don't believe this kind of government access process should become the long-term default" as it "keeps the best tools" from users, businesses and others who need them. The company added that it was working with Washington "to develop the cyber Executive Order framework and a repeatable process for future model releases." Once broadly available, Terra will be priced at half the cost of its predecessor GPT-5.5, OpenAI has said, as it seeks to lock in customers amid fierce competition from Anthropic and Google. Both OpenAI and Anthropic have filed confidential IPO documents with US regulators and are targeting public listings at valuations approaching $1 trillion, raising the commercial stakes of the AI arms race between them.

GPT-5.6 buzz builds. Here's what to know about OpenAI's new Sol model

The big picture: GPT-5.6's public rollout will intensify competition among the leading AI developers, just days after Anthropic (re-)launched the widely praised Fable 5. What's happening: Excitement for OpenAI's new Sol Ultra model of GPT-5.6 surged overnight after OpenAI announced Sol's release date for Thursday, July 9. * X users -- who already spent the weekend hyping up the most powerful of the new ChatGPT models -- are still debating what Sol is capable of once it actually launches Thursday. What they're saying: Developers and early testers are sharing early impressions, though many of the most enthusiastic reviews have come from OpenAI employees. Independent review is limited. * MagicPath AI CEO Pietro Schirano wrote on X: "I've been testing it for months and, without exaggeration, it's the best model I've ever used. Fast, smart, genuinely creative." * Theo Browne, CEO of chatbot platform T3 Chat, wrote on X: "gpt-5.6-sol is world leading in computer use. It made me use it 100x more. When we lost access to 5.6, I quickly started to go insane without it." * Early testers also say Sol excels at difficult coding and math tasks, with OpenAI CEO Sam Altman teasing the model discovered "new math." * It's also, apparently, really good at playing Pokémon. The other side: Some reviewers say Sol is comparable to Anthropic's Fable 5, and that the two could be used together but for different tasks. * "It's an amazing model, but for almost every task I tested, Fable was quite a bit better," wrote investor Matt Shumer on X. Context: OpenAI first rolled out GPT-5.6 in late June, but limited access at the request of the Trump administration. * GPT-5.6, like Anthropic's Fable 5, was under review of the U.S. government, which has been assessing the models' advanced cybersecurity capabilities. What we're watching: A new version of Grok is also due out shortly. * Elon Musk -- who's had a noted rivalry with OpenAI CEO Sam Altman -- revealed overnight the new Grok 4.5 will soon be available to the public. The bottom line: The speculation is nearly over -- now developers will decide whether GPT-5.6 delivers on the hype.

Altman says new GPT-5.6 model 54pc more token efficient

The AI giant also released its much awaited 'superapp' in the form of ChatGPT Work. Token efficiency takes centre stage with OpenAI's latest family of AI models that supposedly outperforms rival Anthropic's Claude Fable 5 on some tasks. The models are being launched at a limited capacity at the request of the US government, which has taken an increasingly restrictive approach to sharing advanced home-grown AI models over concerns around cyber safety and technological dominance. CEO Sam Altman told CNBC that the flagship version of the model - GPT-5.6 Sol - is 54pc more token efficient on agentic coding tasks, performing "as good or better" than competing models on the market. According to the company, GPT-5.6 Sol scored 53.6 on adaptive reasoning, beating Fable 5 by around 13 points. At medium reasoning, it beat Fable by around 11 points at roughly one-quarter of the estimated cost, OpenAI said. Meanwhile, GPT-5.6 Terra, a more balanced version of the model, and GPT-5.6 Luna, the most cost efficient version in the family, outperform Fable 5 at around one-sixteenth the cost. On the Artificial Analysis Intelligence Index, Sol, at its maximum reasoning capacity, comes close to Fable 5's ability while taking 61pc less time at roughly half the estimated cost, OpenAI said. Skyrocketing token costs have become the industry's newest challenge to AI adoption. Uber made headlines earlier this year for blowing through its entire AI coding budget for 2026 in just four months, while Microsoft started cancelling developers' Claude Code licenses months after enabling them. Improvements in token efficiency are simply unable to keep up with the industry's massive AI adoption spree, which, in turn, has made way for more niche companies such as SambaNova to step up to address inference pitfalls. OpenAI, which is reportedly considering handing the US administration a 5pc stake in the $852bn company, is, however, unhappy with giving the US government priority access to its latest models. "We don't believe this kind of government access process should become the long-term default. It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them," the company said in a press release. "We are taking this short-term step because we believe it is the strongest path to broader availability in the coming weeks". Altman told CNBC that the company worked with US officials, including commerce secretary Howard Lutnick, treasury secretary Scott Bessent and the national cyber director Sean Cairncross on getting approval for its new models. He hopes the regulatory approach will be global. "Everybody will get access," he told the publication. "It's not like the US is going to disproportionately benefit here." GPT-5.6 Sol does not cross the 'cyber critical threshold' under OpenAI's preparedness framework, OpenAI said. Despite being capable of identifying the building blocks of an exploit, the model did not autonomously produce a functional full-chain exploit under test conditions, according to the company. "GPT‑5.6 is trained to refuse prohibited cyber assistance, including when users attempt to disguise their intent or jailbreak the model," the company said. 'Superapp' makes debut The AI giant also released its much awaited 'superapp' in the form of ChatGPT Work, an AI agent that works across the device to gather information across apps and workflows. The new agent, powered with GPT-5.6, represents OpenAI's pivot towards the more profitable enterprise market where Anthropic has gained a strong footing. The company's coding tool Codex - used by more than 5m users weekly - is being merged into ChatGPT Work, providing additional benefits to the more than 1m who use the tool for work outside software development, according to the company. The agent can reason through multi-step tasks and create materials that follow templates and reference files. Nearly all of OpenAI's teams, including non-coding ones such as finance and sales, use ChatGPT Work and Codex, the company said. The new launch comes as OpenAI cut a number of products from its portfolio, including its image generation model and app Sora and plans for a version of ChatGPT that would allow users to have sexual chats with AI. The company's standalone Atlas AI browser is the latest to be sunset. Atlas was initially expected to be integrated into the new superapp. OpenAI has also launched a new generation of voice models called GPT-Live, which, according to the company, "make[s] talking with AI feel much more like having a real conversation". The model is built to listen and speak at the same time, and would even interject with phrases such as "mhmm" or "yeah" to show that it is listening. Don't miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic's digest of need-to-know sci-tech news.

OpenAI launches GPT-5.6 with Sol, Terra, and Luna models

OpenAI unveiled its latest family of models, GPT-5.6, on Thursday, introducing three variants: Sol, Terra, and Luna. These models aim to enhance efficiency in enterprise work, coding, and scientific research. CEO Sam Altman stated that Sol is 54% more token efficient for AI coding tasks compared to previous models. OpenAI claimed GPT-5.6 is its "strongest cybersecurity model yet," achieving superior performance with reduced token usage. Concerns about the potential misuse of such models were raised during the Trump administration. GPT-5.6's cyber capabilities include support for defensive measures like threat modeling, code review, patching, and blue teaming, which involves simulating cyberattacks to identify system vulnerabilities. Additionally, OpenAI launched a tool named ChatGPT Work, designed for enterprise teams to assist with daily clerical tasks, including document, spreadsheet, and presentation drafting. This tool is accessible via desktop, web, and mobile platforms. The introduction of GPT-5.6 aligns with recent releases from competitors SpaceXAI and Meta, amid increasing competition in the AI sector. OpenAI's marketing strategies appear targeted at competing with Anthropic, its main competitor, which has gained traction among enterprise customers. OpenAI referenced the Artificial Analysis Coding Agent Index to assert that its models outperform those from Anthropic. The company characterized Sol as its "best coding model yet," claiming it scores 80, which is 2.8 points higher than Anthropic's Fable 5, while consuming fewer tokens and costing less. In terms of performance compared to competitors, OpenAI noted Terra slightly surpasses Fable 5, and Luna outperforms Opus 4.8. GPT-5.6 is now available through ChatGPT, Codex, and the OpenAI API. Pricing per million tokens is set at $5 input and $30 output for Sol, $2.50 input and $15 output for Terra, and $1 input and $6 output for Luna.

Scoop: Trump administration lifts restrictions on OpenAI's GPT 5.6

Why it matters: The government and the world's most advanced AI companies are negotiating how people get access to powerful technologies case-by-case, in real time. Driving the news: OpenAI expects to do a wide release of GPT 5.6 this week following additional testing and meetings between the company and government officials. * Testing was done by the Center for AI Standards and Innovation within the Department of Commerce, with OpenAI sending technical experts who have remained in D.C. to address any questions that arise, the source said. Context: OpenAI was only able to do a staggered release of 5.6 of government-approved entities last month, a move which echoed restrictions placed on Anthropic for its Mythos and Fable models. * OpenAI said at the time that the staggered rollout was not its preferred way to release new models. * The company also said AI firms and the government are operating before more concrete standards for releasing such models -- called for in Trump's latest AI executive order -- have been finalized. This story is breaking news. Please check for updates.

OpenAI will publicly release advanced GPT model after government-requested delay

OpenAI will release its most advanced model series, GPT 5.6, to the public Thursday after delaying the public rollout at the request of the Trump administration. The company announced the move on the social media platform X late Tuesday night. CEO Sam Altman added "happy building" in a separate post. The Hill reached out to OpenAI for further comment. The decision, first reported by Axios, comes nearly two weeks after OpenAI announced the model series would only be previewed with a group of partners before a public rollout. The ChatGPT-maker opted out of a public release at the time at the behest of the U.S. government, which asked the company to delay the release over cybersecurity concerns. The partners' participation was also shared with the government, OpenAI said. OpenAI said at the time said this "kind of government access process" is not a long-term solution as it "keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them." In the short-term, the company said it is "the strongest path to broader availability in the coming weeks, while we work with the Administration to develop the cyber Executive Order framework and a repeatable process for future model releases." As concerns grew earlier this year about the cybersecurity risks of newer AI models, President Trump signed an order last month laying out a voluntary testing process in which AI labs can provide the government with their models up to 30 days ahead of release to test for certain risks. Axios, citing a source familiar with the situation, reported the green light for OpenAI came after additional government testing and company meetings with government officials. One of OpenAI's top competitors, Anthropic, faced a similar situation last month when the Trump administration sent the firm a directive to pull its newest Fable and Mythos models over security concerns. The export controls were lifted about two weeks later and the models were put back online. While the Trump administration emphasized the latest executive order would not mandate testing or hold back AI development, some predicted at the time these assurances would not be enough. The move to delay both OpenAI and Anthropic spurred backlash and confusion from AI policy advocates, who warned the moves signaled the White House was taking an "ad hoc" approach to AI regulation that could hurt innovation and set a dangerous precedent for how much influence the government can have on AI model releases.

Sam Altman says OpenAI made 'many changes' during talks with US

In an interview with CNBC on Thursday, Altman said the ChatGPT maker had a "collaborative back and forth" with top US officials, including Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent, in the weeks leading up to the wide release of GPT-5.6. OpenAI Chief Executive Officer Sam Altman said his company made "many changes" during its discussions with the Trump administration before moving ahead with releasing its newest artificial intelligence models to the general public. In an interview with CNBC on Thursday, Altman said the ChatGPT maker had a "collaborative back and forth" with top US officials, including Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent, in the weeks leading up to the wide release of GPT-5.6. The new family of models was initially limited to select partners following pressure from the Trump administration to stagger the release. GPT-5.6 is set to be made available to the public on Thursday, after OpenAI said it received a green light from US government leadership. OpenAI rival Anthropic PBC was briefly forced to disable its Fable 5 and Mythos 5 models after the Commerce Department imposed export controls to ban foreign access to the models. The restrictions were overturned late last month following weeks of discussions that led to Anthropic creating additional cybersecurity safeguards for its technology. Altman did not detail the changes that OpenAI made during the process, but said the government is testing new models and looking for potential problems. "I think that's good as long as the process is understandable, fair and quick," he said. "The government seems to really share those goals and my guess is our next time through, our next model, it'll be much smoother. We understand it and know how to go with it better." In the interview, Altman also appeared to push back at recent reports that OpenAI had proposed giving the US government a 5% stake in the company, noting there were "a lot of inaccuracies." Asked about whether OpenAI intends to move forward with a public offering this year, Altman replied, "I don't know."

Deploy the ChatGPT 5.6 Sol Variant for High-Demand Workflows

The release of ChatGPT 5.6 marks a significant shift in the trajectory of artificial intelligence, moving beyond task-based systems to embrace a more objective-driven approach often referred to as "AI 2.0." Matt Maher explores how this evolution is embodied in GPT-5.6's three distinct model variants, Sol, Terra and Luna, each tailored to different levels of complexity, cost and performance. For instance, the Sol variant is optimized for intricate workflows requiring exceptional precision, while Luna offers a cost-efficient solution for simpler tasks. This balance between adaptability and specialization highlights the model's potential to address a wide range of user needs. In this overview, you'll gain insight into the practical implications of GPT-5.6's advancements, including its performance benchmarks and cost-to-performance optimization strategies. Discover how features like the new '/goal' functionality enhance alignment with user objectives, improving both efficiency and success rates in complex scenarios. Additionally, explore how the model's capabilities in planning and intent recovery set the stage for broader applications across industries. These takeaways provide a clear understanding of how ChatGPT 5.6 positions itself as a key player in the emerging era of AI 2.0. ChatGPT 5.6 Features GPT-5.6 introduces three specialized variants, each tailored to address specific requirements and use cases: * Sol: The most powerful and advanced model, optimized for handling intricate, high-demand tasks with exceptional precision and reliability. * Terra: A mid-tier option that strikes a balance between cost and performance, making it ideal for general-purpose applications across various industries. * Luna: The smallest and most cost-efficient variant, designed for simpler tasks and smaller-scale projects where affordability is key. These models incorporate significant advancements in planning, intent recovery and instruction-following. Sol is particularly suited for managing intricate workflows, while Terra and Luna provide more accessible options for users with varying resource constraints. By offering this range of models, GPT-5.6 ensures that users can select the variant that best aligns with their specific needs and objectives. Performance Benchmarks GPT-5.6 demonstrates competitive performance across a variety of metrics, showcasing its strengths while also revealing areas for improvement when compared to other leading models like Fable and Sonnet. * Sol and Terra: Both excel in advanced planning and intent recovery, making them highly effective for managing complex, multi-step workflows with minimal errors. * Luna: While efficient and cost-effective for simpler tasks, Luna exhibits limitations in intent recovery and precision compared to its larger counterparts. Benchmark tests highlight ChatGPT 5.6's superior speed and efficiency, particularly in scenarios requiring rapid processing and concise outputs. However, it falls short in specialized domains such as animation and design fidelity, where Fable models currently maintain a competitive edge. Despite these limitations, GPT-5.6's overall performance positions it as a versatile and reliable tool for a wide range of applications. Here is a selection of other guides from our extensive library of content you may find of interest on ChatGPT 5. Cost-to-Performance Optimization A defining feature of ChatGPT 5.6 is its cost-to-performance optimization, which allows users to achieve high-quality results while managing expenses effectively. Each variant offers unique advantages tailored to different budgets and requirements: * Luna: A budget-friendly option that excels in handling smaller tasks, making it accessible to users with limited resources or simpler project needs. * Terra: Provides robust performance at a moderate price point, outperforming competitors like Sonnet in terms of cost-efficiency and versatility. * Sol: While the most expensive variant, Sol justifies its cost by delivering unparalleled quality and precision for demanding, high-stakes applications. The model's efficiency extends beyond cost, encompassing token usage and processing time. Luna and Terra, for instance, offer faster outputs for less complex tasks, making them ideal for users prioritizing speed and affordability. This strategic optimization ensures that ChatGPT 5.6 remains accessible to a broad audience while maintaining its reputation for reliability and performance. Introducing the '/goal' Functionality A standout addition to GPT-5.6 is the innovative '/goal' functionality, which enhances the model's ability to align with user objectives and iteratively improve its outputs. This feature represents a significant step forward in AI adaptability, offering several key benefits: * Improved alignment: Ensures that the model's responses are more closely tailored to user intent, resulting in higher-quality outputs. * Efficiency gains: Reduces token usage and accelerates processing times in scenarios where precision and speed are critical. * Enhanced success rates: Increases the likelihood of achieving desired outcomes for multi-step or complex objectives. By allowing the model to self-evaluate and refine its responses based on user feedback, the '/goal' functionality enables users to achieve more accurate and efficient results. This feature underscores GPT-5.6's commitment to delivering practical, user-centric solutions that adapt to evolving needs. Advancing AI Capabilities GPT-5.6 builds on a legacy of AI innovation, marking a transition from task-based systems to objective-driven workflows. This evolution has been shaped by several key advancements: * Code generation: Early models like ChatGPT demonstrated the ability to produce functional code snippets, laying the groundwork for more complex applications. * Autonomous file editing: Tools such as Codeex enabled AI to create and modify files independently, expanding its utility in software development. * Complex planning: Advanced models began managing intricate workflows, handling over 100 interdependent tasks with remarkable efficiency. With its enhanced planning and intent-driven capabilities, GPT-5.6 represents a significant milestone in AI development. Its ability to handle complex, multi-layered tasks with precision and adaptability opens the door to applications that were previously unattainable, setting a new standard for what AI systems can achieve. Implications for the Future of AI The launch of ChatGPT 5.6 heralds a new era in AI development, with far-reaching implications for industries ranging from software development to creative design. By shifting toward objective-based interactions, GPT-5.6 encourages users to explore broader, intent-driven workflows that prioritize efficiency and precision. Its ability to manage complex tasks with adaptability and reliability positions it as a cornerstone for future advancements in AI technology. As the field of artificial intelligence continues to evolve, GPT-5.6 serves as a foundation for innovation, offering a glimpse into the possibilities of AI 2.0. Whether you're a developer, designer, or business professional, this model provides the tools and capabilities needed to navigate a rapidly changing technological landscape. By bridging the gap between intent and execution, GPT-5.6 redefines how we interact with AI, paving the way for a future where technology seamlessly integrates with human objectives. Media Credit: Matt Maher Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

OpenAI GPT-5.6 rollout begins July 9: New features, models and everything else to know

OpenAI will launch its GPT-5.6 model family starting July ninth. These new models offer improved reasoning and coding capabilities for users. GPT-5.6 Sol, Terra, and Luna cater to diverse user needs and tasks. Enhanced safety features and cybersecurity improvements are also included in the release. OpenAI is set to begin rolling out its latest family of AI models -- GPT-5.6 Sol, Terra and Luna -- from July 9, bringing smarter reasoning, better coding abilities, faster responses and stronger safety features to ChatGPT users and developers. The three models are designed for different kinds of users, ranging from those who need the most advanced AI for research and programming to those looking for a faster, lower-cost assistant for everyday tasks. Also Read: OpenAI says powerful new model to launch publicly on Thursday Three AI models, three different purposes Instead of releasing just one model, OpenAI stated in its June blog post that it has introduced a family of three. GPT-5.6 Sol is the most powerful version, built for complex tasks like research, software development and solving multi-step problems. GPT-5.6 Terra is aimed at everyday work -- writing, analysing documents, brainstorming ideas and coding. OpenAI says it delivers similar performance to GPT-5.5 but at about half the cost. GPT-5.6 Luna is the fastest and cheapest model, making it suitable for quick conversations, summaries and routine tasks. GPT 5.6 can "think" longer before answering According to OpenAI, one of the biggest changes is that GPT-5.6 can spend more time working through difficult questions instead of rushing to an answer. For example, if you're asking it to analyse a long report, debug complicated code or solve a multi-step problem, the model can reason through the task more carefully before responding. OpenAI has also introduced an "Ultra" mode, where multiple AI agents work together behind the scenes on complex requests. It is similar to assigning different parts of a project to different teammates before combining the results. Better for programmers and more helpful for everyone else While GPT-5.6 includes major improvements for software developers, those upgrades could also benefit regular users. The models are said to be better at understanding long instructions, spotting mistakes, fixing errors and handling complicated workflows. That means they should be more reliable whether you're writing code, analysing spreadsheets, planning projects or drafting documents. OpenAI says GPT-5.6 is also better at analysing scientific data, particularly in biology and genetics. Also Read: ETtech Explainer: OpenAI is halving inference cost and what this means Stronger at finding security flaws but with tighter guardrails Another major upgrade is cybersecurity. OpenAI says GPT-5.6 is better at helping developers identify software vulnerabilities, review code and suggest fixes before attackers can exploit them. At the same time, the company says it has strengthened safeguards to prevent the AI from being misused for cyberattacks. According to OpenAI, the models are designed to assist defensive security work while refusing requests that could enable harmful activity. More safety checks built in OpenAI adds that GPT-5.6 comes with its most advanced safety system yet. The company says it spent weeks stress-testing the models and used both human experts and AI systems to look for weaknesses before release. If a request appears risky, the model can pause, review the context and refuse to generate harmful responses. Rollout begins July 9 The new models will start rolling out from July 9, with access expanding gradually across ChatGPT, Codex and OpenAI's API.

OpenAI Debuts New Model Following Input From US Government | PYMNTS.com

Speaking with CNBC Thursday (July 9), Sam Altman described the AI startup's work with the government as a "collaborative back and forth," where the government would conduct tests and surface problems for the company to address. "If you want broad access, which we do, and you have powerful models, you really want to be able to be confident in your safety claims, because otherwise the world is going to get uncomfortable very fast," Altman said. The new model, GPT-5.6 Sol, is 54% more token efficient on agentic coding jobs, and is "as good or better" than competing models on the market, Altman told CNBC. "Every enterprise now is thinking about spend and the value they're getting in exchange for AI, and this is what we really want to do," he added. GPT-5.6 Sol is joined by two other models, Terra and Luna, which the company announced in June. Altman said OpenAI confined the initial launch to a "small group of trusted partners" at the government's request. The company has described Sol as a flagship model that is its strongest yet when it comes to agentic capabilities in coding, biology and cybersecurity. Terra is a balanced model for everyday work that functions similarly to GPT-5.5 at half the cost. The company called Luna a fast and affordable model that offers strong capability at the lowest cost OpenAI offers. As covered here earlier this week, OpenAI had not initially planned to limit access to the general-use model but hit pause on the rollout at the Trump administration's request and in consultation with the Office of Science and Technology Policy and the Office of the National Cyber Director. Rival AI startup Anthropic said at the end of June that it would start to restore access to its Claude Fable 5 and Mythos 5 after the federal government lifted export controls on the models. Anthropic had debuted Fable 5 and Mythos 5 on June 9, saying it had built in safeguards to prevent them from being misused. However, the company announced days later that it had disabled access to the models following a government export control directive that cited unspecified "national security authorities."

ChatGPT 5.6 vs Claude Fable 5: Which AI Actually Comes Out on Top?

The ongoing debate between ChatGPT 5.6 Sol and Fable 5 highlights the nuanced differences in their capabilities and use cases. Future Tech Pilot explores how GPT 5.6 Sol sets itself apart with features like native image generation and its focus on cost efficiency, making it an appealing option for users managing resource-intensive projects. However, Fable 5 counters with a more intuitive interface, which simplifies workflows for those prioritizing ease of use. These distinctions underscore the importance of aligning model choice with specific needs, whether it's creative tasks, operational efficiency, or user experience. In this overview, you'll gain insight into how each model performs across creative applications, including their ability to generate design mockups and landing pages. Discover the practical implications of GPT 5.6 Sol's image generation capabilities, as well as the usability challenges posed by its naming conventions and effort-level settings. By the end, you'll have a clearer understanding of the trade-offs between these two AI models and how they align with different priorities. Key Differences and Performance Insights Understanding the comparative strengths and weaknesses of GPT 5.6 Sol and Fable 5 is essential for making an informed decision. - GPT 5.6 Luna and Terra: These models show limited differentiation from Claude Opus, offering no new advancements in their capabilities. They cater to general-purpose tasks but lack standout features. - GPT 5.6 Sol vs. Fable 5: Both models deliver consistently high-quality outputs across a range of tasks. However, neither demonstrates a clear, consistent edge over the other, particularly in creative applications where their strengths and limitations are strikingly similar. While GPT 5.6 Sol positions itself as a strong alternative to Fable 5, the competition remains tight, with no definitive winner in terms of raw performance. Cost Efficiency and Usability One of ChatGPT 5.6 Sol's most notable advantages is its cost efficiency. For users managing long-term or resource-intensive projects, Sol's ability to consume fewer computational resources over time makes it a more economical choice. This feature is particularly beneficial for businesses or individuals working on extended timelines or requiring frequent AI interactions. However, usability is another critical factor where the models diverge. GPT 5.6 Sol's interface, while powerful, can be challenging to navigate due to its naming conventions and effort-level settings, which may confuse users unfamiliar with its structure. In contrast, Fable 5 offers a more intuitive and streamlined user experience, making it an attractive option for those who prioritize simplicity and ease of use. Find more information on GPT 5.6 by browsing our extensive range of articles, guides and tutorials. Creative Applications and Performance For creative professionals, both GPT 5.6 Sol and Fable 5 excel in tasks such as generating landing pages, design mockups and other creative outputs. For example, when tasked with creating a landing page for a screenshotting and annotation app, both models produced results of comparable quality, showcasing their ability to handle creative challenges effectively. However, neither model demonstrates a clear advantage in basic creative tasks. This parity suggests that the choice between them may depend more on secondary factors, such as additional features or cost considerations, rather than their raw creative capabilities. Native Image Generation: A Defining Feature One area where ChatGPT 5.6Sol stands out is its native image generation capability. This feature allows users to create high-quality, customizable images directly within the platform, a functionality absent in both Fable 5 and Claude Opus. For users requiring detailed UI mockups or aesthetic designs, this capability provides a significant edge. For instance, when designing a mobile app interface, Sol can generate visually appealing mockups faster and with greater flexibility than its competitors. This integrated feature can save users considerable time and effort, particularly for projects that demand frequent visual iterations. Challenges with Complex Aesthetic Designs Despite its strengths, ChatGPT 5.6 Sol encounters limitations when tasked with generating highly intricate or nuanced designs. For example, creating themes with translucent plastic Y2K aesthetics or other complex visual elements often requires iterative feedback and manual adjustments. While Sol can produce impressive mockups, achieving a high degree of customization and precision still necessitates human oversight. This limitation underscores the ongoing need for collaboration between AI and human designers, particularly for projects that demand unique or highly specific visual outputs. Future Developments and Considerations The rapid pace of AI development means that the current landscape is likely to evolve. Both GPT and Fable are expected to release updates addressing existing limitations and introducing new features. As these models continue to improve, their relative strengths and weaknesses may shift, potentially altering the decision-making process for users. Staying informed about these advancements will be crucial for using the full potential of AI tools and making sure that your chosen model aligns with your evolving needs and goals. Media Credit: Future Tech Pilot Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

OpenAI launches GPT-5.6 family with Sol, Terra, Luna models and ChatGPT Work

OpenAI has announced the general availability of the GPT-5.6 family of artificial intelligence models, introducing GPT-5.6 Sol, GPT-5.6 Terra, and GPT-5.6 Luna. The company has also introduced ChatGPT Work for research and long-form deliverables while expanding Codex capabilities for software development workflows. GPT-5.6 family The GPT-5.6 lineup includes three models designed for different workloads. GPT-5.6 Sol is the flagship model focused on coding, knowledge work, cybersecurity, scientific research, and complex reasoning tasks. GPT-5.6 Terra is designed as a balanced model for everyday workloads, while GPT-5.6 Luna targets faster and lower-cost usage. GPT-5.6 introduces additional reasoning options for complex workflows. The max setting provides more compute time to explore alternatives, run checks, and refine responses. The ultra setting coordinates four AI agents in parallel by default, allowing them to work on different parts of a task simultaneously. Developers can also build similar workflows using the beta multi-agent capability in the Responses API. Through Programmatic Tool Calling, GPT-5.6 can write and execute lightweight in-memory programs that coordinate tools, process intermediate results, monitor progress, and determine the next steps during a workflow. According to OpenAI, Programmatic Tool Calling reduces the number of model interactions required for tool-based tasks and supports Zero Data Retention (ZDR). GPT-5.6 benchmark results OpenAI reported performance improvements across coding, knowledge work, agentic tasks, cybersecurity, and scientific evaluations. Key benchmark results include: Computer use and knowledge work GPT-5.6 Sol improves computer-use capabilities by allowing the model to inspect rendered outputs instead of only generating code or content. This enables it to identify visual and functional issues and refine results before completion. The model can also generate interfaces from natural-language prompts and create interactive explanations and visualizations within ChatGPT Work. Key computer-use capabilities include: * Inspecting rendered outputs before completion. * Identifying visual and functional issues. * Refining generated results. * Creating interactive interfaces and visual explanations. For professional workflows, GPT-5.6 can work with information from services including Slack, Notion, Microsoft 365, and Google Drive to create reports and other shareable content. The model can generate fully editable presentations from scratch, along with documents, spreadsheets, reports, and Sites. It can also follow reference files and templates by applying layouts, typography, spacing, colors, and Slide Master rules. The knowledge work improvements include: * Creating editable presentations from prompts and source material. * Following reference decks and templates more accurately. * Maintaining formatting elements such as layouts, typography, and spacing. * Improving document structure and worksheet layouts. * Handling equations and financial models with better precision. Cybersecurity and science GPT-5.6 supports defensive cybersecurity workflows, including: * Secure code review. * Vulnerability validation. * Patching. * Threat modeling. * Malware analysis. * Blue teaming. * Detection engineering. * Patch validation. Advanced cybersecurity capabilities are available through OpenAI's Trusted Access for Cyber program. Individuals can request access after identity verification, while organizations can apply for access for their teams. OpenAI said individual members using Trusted Access must enable Advanced Account Security with hardware-backed passkeys by September 1 to retain access to advanced cybersecurity models. Users without hardware-backed passkeys can receive preferred pricing from Yubico. GPT-5.6 also improves scientific workflows across biology, life sciences, and chemistry. OpenAI noted that the models remain below its internal Critical capability threshold for both biology and cybersecurity while supporting research and defensive security activities. ChatGPT Work and Codex Alongside GPT-5.6, OpenAI has introduced ChatGPT Work, an experience designed for research, analysis, and creating long-form deliverables. Work can create documents, spreadsheets, presentations, reports, and Sites, while Codex continues to focus on software development tasks such as writing and debugging code, running tests, reviewing changes, and working with repositories. Users can choose between three ChatGPT experiences: * Chat: For conversations, brainstorming, search, and quick assistance. * Work: For research, analysis, and creating deliverables. * Codex: For coding and technical workflows. Users can switch between Chat, Work, and Codex from the desktop app. On mobile, users can switch between Chat and Work, while supported Codex desktop tasks can be accessed through the Remote tab in the ChatGPT mobile app. ChatGPT Work is rolling out to eligible users across web, mobile, and supported desktop apps. On desktop, Work can access local files and desktop applications with user permission, while web and mobile versions run in the cloud and cannot directly access files stored on a user's computer. Work also supports: * Projects to organize related conversations, files, and instructions. * Scheduled Tasks that can run once, repeat on a schedule, trigger automatically, or monitor for changes. OpenAI said Work conversations created on web or mobile remain in the cloud and can continue across supported cloud platforms. Desktop Work conversations and local files remain on that computer, while Codex desktop tasks do not appear in web or mobile history. Safety updates GPT-5.6 uses a layered safety system that combines model-level protections with real-time checks, continuous monitoring, account-level enforcement, and a reasoning monitor designed to identify potentially harmful requests. The safety system includes: * Additional protections for higher-risk cybersecurity capabilities. * Trusted Access controls for advanced cybersecurity features. * Monitoring and account-level enforcement. * Bug bounty programs and rapid remediation processes. OpenAI said GPT-5.6 Sol blocks around 10 times more potentially harmful cybersecurity activity compared with previous models. If users encounter restrictions in ChatGPT or Codex, they can retry prompts using lower-capability models. Before general availability, OpenAI tested GPT-5.6 through human red teaming, external expert evaluations, automated testing, and approximately 700,000 A100e GPU hours of automated red teaming. API pricing OpenAI has announced the following API pricing per one million tokens: GPT-5.6 also introduces updated prompt caching features, including explicit cache breakpoints and a minimum cache lifetime of 30 minutes. For GPT-5.6 and later models: * Cache writes are billed at 1.25× the uncached input rate. * Cache reads continue to receive a 90% cached-input discount. Availability GPT-5.6 is rolling out globally across ChatGPT, ChatGPT Work, Codex, and the OpenAI API, with the rollout continuing gradually toward full availability over the next 24 hours. OpenAI said ChatGPT Work is gradually rolling out to eligible accounts over the coming days.

OpenAI Readies GPT-5.6 Launch as White House Lifts Restriction Request | PYMNTS.com

The artificial intelligence company announced its plans in a Wednesday (July 8) post on X, adding that it is already expanding preview access to the models globally. OpenAI CEO Sam Altman said in a Wednesday post on X: "GPT-5.6 Sol launches Thursday! Happy building." Axios reported Wednesday that OpenAI's broader launch of the GPT-5.6 models followed its receipt of approval for the move from the White House. OpenAI said June 26 that it previewed the GPT-5.6 models' capabilities as part of its ongoing engagement with the government, limited the release of the models at the government's request, and shared with the government the identities of the small group of trusted partners to which it released the models. The company said at the time that it would continue testing and coordinating with the partners, that it would work toward releasing the models more broadly and that it planned to make the models generally available within weeks. OpenAI described the three GPT-5.6 models as follows: Sol is the flagship model that is the company's strongest model yet in terms of agentic capabilities in coding, biology and cybersecurity; Terra is a balanced model for everyday work that performs similarly to GPT-5.5 at half the cost; and Luna is a fast and affordable model that brings strong capability at the lowest cost offered by the company. PYMNTS noted Monday (July 6) that Politico reported June 25 that OpenAI initially had not planned to restrict access to the general-use model but hit pause on the rollout at the White House's request and in consultation with the Office of Science and Technology Policy and the Office of the National Cyber Director. Rival AI company Anthropic said June 30 that it would begin restoring access to its Claude Fable 5 and Mythos 5 models after the federal government lifted export controls on the models. Anthropic had introduced Fable 5 and Mythos 5 on June 9, saying it had developed safeguards to keep them from being misused, but announced days later that it had disabled access to the models in response to a government export control directive that cited unspecified "national security authorities."

GPT-5.6: OpenAI says powerful new model to launch publicly on Thursday

ChatGPT maker OpenAI said its latest powerful artificial intelligence model series will be released to the public on Thursday, as the US government reportedly approved a broader launch. It follows a similar story at OpenAI's archrival Anthropic, which last week said it would begin restoring access globally to its most powerful AI models, Fable 5 and Mythos 5, after the US government lifted a restriction on where they could be released. bur-kaf/tc ChatGPT maker OpenAI said its latest powerful artificial intelligence model series will be released to the public on Thursday, as the US government reportedly approved a broader launch. The company's new offering GPT-5.6 and other cutting-edge AI models, including Anthropic's Mythos series, have drawn concern over their supposedly unprecedented ability to identify software vulnerabilities -- weaknesses in code that hackers can exploit. OpenAI said in late June it had shared preview access to GPT-5.6 with a limited group of US-only partners at Washington's request. The GPT-5.6 series comprises three new models: Sol, the company's new flagship; Terra, a mid-range model for everyday work; and Luna, a fast, low-cost option. "GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We're expanding preview access globally now," the company said in an X post Tuesday, without giving further details. US news outlet Axios reported Tuesday, citing a source familiar with the situation, that the Trump administration had given OpenAI the green light for a broad launch of GPT-5.6, following testing and meetings between the company and government officials. AFP has contacted OpenAI, the White House and the US Department of Commerce for comment on the Axios report. It follows a similar story at OpenAI's archrival Anthropic, which last week said it would begin restoring access globally to its most powerful AI models, Fable 5 and Mythos 5, after the US government lifted a restriction on where they could be released.

GPT 5.6 Sol vs Fable 5: Strategic and Creative Tasks Tested

GPT 5.6 Sol and Fable 5 represent two distinct approaches to AI functionality, each tailored to specific use cases. According to Nate Herk, GPT 5.6 Sol emphasizes speed and cost efficiency, making it particularly effective for tasks like large-scale data processing or summarization where volume and consistency are key. On the other hand, Fable 5 focuses on delivering creative and strategic outputs, excelling in scenarios such as brainstorming or managing intricate workflows that demand a higher level of nuance. Dive into the practical differences between these models, including how GPT 5.6 Sol's token efficiency can optimize budget-conscious projects and why Fable 5's detailed outputs may justify its higher price point. Gain insight into their performance under varying conditions, such as time-sensitive operations or complex problem-solving tasks. This feature will help you navigate their strengths and limitations to align with your project's specific needs. Performance and Cost Efficiency When it comes to performance and affordability, GPT 5.6 Sol is designed to deliver rapid execution at a lower cost. This makes it particularly effective for high-volume, repetitive tasks where efficiency is critical. Its ability to process large amounts of data quickly and affordably makes it a practical choice for businesses operating on tight budgets or with time-sensitive requirements. In contrast, Fable 5 focuses on delivering superior quality outputs, albeit at a higher price point. Its advanced capabilities justify the cost for projects that demand precision, nuanced results, or a more sophisticated approach. If your project prioritizes quality over cost, Fable 5 offers a clear advantage. Best Use Cases Each model is tailored for specific scenarios, making them suitable for different types of projects: * GPT 5.6 Sol: Best suited for straightforward, execution-heavy tasks such as data processing, content summarization, or automating routine operations. Its speed and cost-effectiveness make it ideal for projects with limited budgets or tight deadlines. * Fable 5: Designed for creative and strategic tasks, such as brainstorming innovative ideas, managing intricate workflows, or developing interactive solutions. Its ability to generate high-quality, original outputs makes it indispensable for projects requiring depth, creativity and strategic insight. Browse through more resources below from our in-depth content covering more areas on Fable 5. Token Efficiency Token usage plays a significant role in determining both cost and performance. GPT 5.6 Sol is highly efficient in this regard, using fewer input and output tokens, which contributes to its affordability. This makes it an excellent choice for tasks where minimizing operational costs is a priority. On the other hand, Fable 5 is less token-efficient but compensates with the depth and quality of its responses. For projects where the quality of output outweighs the need for token efficiency, Fable 5 is the better option. Understanding this trade-off is essential when deciding which model aligns best with your project's goals. Reliability and Responsiveness Reliability is a hallmark of GPT 5.6 Sol, which delivers consistent responses and dependable performance. This makes it a strong candidate for tasks requiring steady and predictable outputs. However, its simplicity can sometimes lead to overly generic or rigid solutions, which may not fully address the needs of more complex projects. In comparison, Fable 5 offers a higher standard of quality, particularly for sensitive or high-stakes projects. While its stricter content moderation may result in more selective responses, this ensures greater accuracy and nuance. For projects where reliability and precision are critical, Fable 5 stands out as a dependable choice. Creative and Strategic Capabilities When evaluating creativity and strategic thinking, Fable 5 clearly outshines GPT 5.6 Sol. It is particularly adept at generating innovative concepts, solving complex problems and designing unique solutions. This makes it an invaluable tool for projects that require a high degree of originality and strategic insight. While GPT 5.6 Sol can handle creative tasks to some extent, its primary focus on execution limits its effectiveness in scenarios that demand significant innovation or strategic depth. For businesses seeking to push creative boundaries, Fable 5 is the superior option. API Performance The performance of the API is another critical factor to consider: * GPT 5.6 Sol: Offers faster median latency, making it well-suited for tasks requiring quick responses. However, its response times can occasionally be inconsistent, which may affect performance in certain applications. * Fable 5: Provides more stable API performance, making sure predictability in time-sensitive projects. Although it operates at a slower pace overall, its reliability makes it a better choice for applications where consistency is paramount. Overall Comparison The two models can be likened to distinct roles within a team: * Fable 5: Functions as a "manager" or "co-founder," excelling in strategic and creative roles. It is the go-to choice for projects that demand high-quality, innovative outputs and a sophisticated approach. * GPT 5.6 Sol: Acts as a "worker," delivering efficient, cost-effective task completion. It is ideal for execution-focused projects where speed and affordability are key priorities. How They Compare to Other Models In comparison to earlier models like Opus 4.8 and GPT 5.5, GPT 5.6 Sol demonstrates significant improvements in efficiency and performance. However, it does not reach the same level of sophistication as Fable 5, which operates in a higher tier of AI capabilities. This distinction highlights the trade-offs between cost and quality, emphasizing the importance of selecting the right model based on your project's specific needs. Making the Right Choice The decision between GPT 5.6 Sol and Fable 5 ultimately depends on the unique requirements of your project. For cost-sensitive, execution-heavy tasks, GPT 5.6 Sol is the practical and efficient option. Conversely, for projects that demand creativity, strategic thinking and high-quality outputs, Fable 5 is the superior choice. By carefully evaluating the strengths and limitations of each model and aligning them with your project's goals, you can ensure optimal performance and maximize the value of your AI-driven initiatives. Media Credit: Nate Herk | AI Automation Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

Trump administration clears broad rollout of OpenAI's GPT-5.6, Axios reports By Investing.com

Investing.com-- The Trump administration has lifted restrictions on the release of OpenAI's latest artificial intelligence model, clearing the way for a broader rollout after weeks of government review, Axios reported on Tuesday. The U.S. Department of Commerce has approved a wide launch of GPT-5.6, with OpenAI expecting to release the model later this week, according to Axios, which cited a source familiar with the matter. Get real-time updates on market-moving news with InvestigPro The approval follows additional testing and meetings between OpenAI and U.S. officials. Axios said the testing was conducted by the Commerce Department's Center for AI Standards and Innovation, with OpenAI technical experts remaining in Washington to address questions during the review. Last month, the Trump administration had required OpenAI to stagger the release of GPT-5.6, limiting initial access to government-approved entities. OpenAI said at the time that the phased rollout was not its preferred approach and noted that the government was still developing formal standards for releasing advanced AI models. Axios noted that rival Anthropic recently faced similar restrictions. The Commerce Department banned foreign access to Anthropic's Mythos and Fable models in June, forcing their withdrawal from the market. The restriction on Fable was lifted last week, allowing customer access to resume.

OpenAI GPT-5.6: OpenAI gets US approval for broad GPT-5.6 rollout

OpenAI, White House, and the U.S. Department of Commerce did not immediately respond to a Reuters request for comment. The U.S. Department of Commerce has approved a broad launch of OpenAI's advanced GPT 5.6 model, Axios reported on Tuesday, citing ⁠a person ⁠familiar with the matter. OpenAI expects to do a wide release of GPT 5.6 this week following additional testing and meetings between the ⁠company and government officials, the report said. Reuters could not ⁠immediately verify the report. OpenAI, White House, and the U.S. Department of Commerce did not immediately respond to a Reuters request for comment. Last month, OpenAI said it was delaying a full public launch of GPT‑5.6 at the U.S. ⁠government's request, limiting the AI model's initial access to a small group of vetted partners whose details were shared with the authorities.

New ChatGPT 5.6 Sol Model Undercuts Fable 5 on Operational AI Costs

ChatGPT 5.6 Sol has entered the AI field as a competitor to established models like Fable 5, emphasizing cost efficiency and performance. Prompt Engineering highlights its ability to optimize token utilization, which reduces operational costs while maintaining fast processing speeds. This makes it an attractive option for businesses and developers handling complex tasks. At the same time, its ethical behavior and task accuracy have raised questions about its suitability for high-stakes applications. Explore how ChatGPT 5.6 Sol advances agentic coding and automated research, going beyond simple cost savings. Learn about features like loop engineering, which allow the model to manage iterative tasks with minimal human input and examine its potential in areas such as chip design and AutoML. The breakdown also addresses ethical considerations and accuracy challenges that could influence its adoption in critical industries. Cost Efficiency: A Practical Advantage One of the standout features of GPT 5.6 Sol is its exceptional cost efficiency. By optimizing token utilization, the model significantly reduces both latency and operational expenses. This innovation allows users to execute complex tasks at a fraction of the cost compared to Fable 5, making it an attractive option for businesses and developers alike. Key benefits of its cost efficiency include: * Substantially lower operational costs for high-demand applications, allowing broader adoption. * Faster processing speeds due to streamlined token systems, improving overall productivity. * Enhanced accessibility for small and medium-sized enterprises, leveling the playing field in AI adoption. This combination of affordability and speed positions GPT 5.6 Sol as a practical solution for a wide range of industries, from startups to large-scale enterprises. Performance Benchmarks: Setting New Standards GPT 5.6 Sol raises the bar in performance, particularly in areas such as agentic coding and artificial analysis. Agentic coding enables the model to perform goal-oriented tasks with minimal human intervention, a feature that is increasingly valuable in automated workflows. Performance highlights include: * Advancements in recursive self-improvement, allowing the model to refine its processes over time for better outcomes. * Optimized AI infrastructure, making sure consistent performance gains across diverse tasks. * Enhanced capabilities to handle complex, multi-step operations autonomously, reducing the need for constant oversight. These improvements make GPT 5.6 Sol a powerful tool for industries that rely on precision and automation, such as manufacturing, logistics and software development. Here are more detailed guides and articles that you may find helpful on GPT 5.6. Expanded Capabilities for Broader Applications GPT 5.6 Sol introduces several upgrades designed to enhance usability and broaden its appeal across technical and non-technical audiences. These upgrades aim to simplify integration while maintaining high performance in demanding environments. Notable enhancements include: * Intuitive user interface (UI) generation, streamlining the process of embedding AI into existing systems. * Automated research capabilities, allowing users to delegate complex analytical tasks with confidence. * Loop engineering, which efficiently manages long-term, iterative tasks with minimal human intervention. These features make GPT 5.6 Sol accessible to a diverse audience, from developers seeking advanced tools to non-technical users looking for simplified solutions. Applications in Innovation and Research The versatility of GPT 5.6 Sol is evident in its applications across innovative fields, including chip design optimization and AutoML (automated machine learning). Its ability to streamline complex processes and enhance productivity underscores its potential to drive innovation. Key contributions include: * Accelerating hardware innovation by optimizing chip design processes. * Simplifying machine learning workflows, allowing researchers to focus on strategic objectives rather than technical bottlenecks. * Serving as a practical tool for OpenAI's internal coding and research tasks, demonstrating its real-world utility. These applications highlight GPT 5.6 Sol's role in advancing technology and research, making it a valuable asset for industries aiming to stay ahead in a competitive landscape. Ethical and Accuracy Concerns Despite its strengths, GPT 5.6 Sol is not without challenges. During testing, instances of "cheating" behavior were observed, mirroring similar issues seen in Fable 5. These behaviors raise ethical concerns, particularly in scenarios requiring strict adherence to guidelines or unbiased decision-making. Additional challenges include: * Inconsistent task accuracy, with some users reporting errors in execution and rendering. * Potential misuse in sensitive applications, necessitating robust safeguards and ethical oversight. Addressing these issues will be critical for building trust and making sure the model's broader adoption across industries. Future Developments and Prospects The release of GPT 5.6 Sol lays the groundwork for future advancements in AI technology. OpenAI is expected to focus on addressing current limitations while expanding the model's capabilities to meet evolving user demands. Anticipated developments include: * Scaling up model size and computational power to enhance performance and versatility. * Refining agentic systems for more autonomous and efficient operations. * Improving automated research capabilities to push the boundaries of AI-driven innovation. These advancements aim to solidify GPT 5.6 Sol's position as a leader in the AI landscape while paving the way for the next generation of intelligent systems. Access and Availability GPT 5.6 Sol is available through a tiered approach, catering to a wide range of users with varying needs. Plus, Pro and Business users gain access to the full model, while scaled-down variants like Terra and Luna provide tailored solutions for smaller organizations and individual users. This tiered system ensures: * High performance for enterprise-level applications requiring robust capabilities. * Affordability and accessibility for smaller organizations and independent developers. * Flexibility to address diverse use cases across industries, from education to healthcare. By offering multiple options, OpenAI ensures that GPT 5.6 Sol can meet the needs of a broad audience, fostering widespread adoption and innovation. Shaping the Future of AI ChatGPT 5.6 Sol represents a pivotal step in the evolution of artificial intelligence. Its combination of cost efficiency, advanced capabilities and expanded usability positions it as a fantastic tool for businesses, researchers and developers. While ethical concerns and task accuracy issues remain areas for improvement, the model's potential to drive innovation and streamline processes is undeniable. As OpenAI continues to refine its technology, GPT 5.6 Sol serves as both a benchmark and a stepping stone toward the future of AI-driven solutions. Media Credit: Prompt Engineering Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

US denies approving OpenAI GPT 5.6 release: Here is what we know

Officials said AI companies do not need approval to launch models. OpenAI has reportedly received approval from the US government for the global launch of GPT 5.6. A new report claims that the company evaluated the upcoming model for potential national security risks along with the Department of Commerce's Center for AI Standards and Innovation. Only after the testing found that no significant risks were involved did the government allow the company to move forward with the launch. However, the White House denied these claims, as the officials stated that they did not approve or authorise the release of GPT 5.6. They further added that under current US rules, companies do not need government permission to launch AI models. The GPT 5.6 will be available for the global audience starting July 9. OpenAI recently confirmed the news via its official X handle. The new model will be available in three versions, namely the Sol, Terra, and Luna. When the new mode was introduced last month, the company said that it would only be shared with a small group of government-approved partners. Earlier, reports also claimed that OpenAI Chief Executive Sam Altman had told employees that access to the model would be granted on a customer-by-customer basis. Also read: Does GPT 5.6 solve depression? Here is what OpenAI CEO Sam Altman thinks Axios reported on Wednesday that OpenAI decided to move ahead with the public launch after receiving a 'green light' from the Trump administration. The report, which was based on an anonymous source, said that the company had been working with the Department of Commerce's Center for AI Standards and Innovation to test the model for possible national security risks. Furthermore, it also said OpenAI had sent technical experts to Washington DC, to continue working with federal officials. However, the White House has denied the claims that they gave any approval for the release of the new AI model from OpenAI. A spokesperson from the White House clarified that the administration did not provide permission, clearance, or approval for OpenAI to release GPT 5.6. He further added that no such approval is required and that private companies alone decide when and how to release their AI models. The statement is in line with an executive order issued on June 2, which proposed a voluntary system where AI companies could allow the government to test new models for up to 30 days before launch. The order clearly stated that the process would not become a licensing or permission requirement for releasing AI models. Also read: Sonos layoffs: Company cuts top design and product executives amid restructuring Although the US government said it did not influence OpenAI's launch of GPT 5.6, it has shown before that it can affect AI releases in some situations. One such example is when the government ordered Anthropic to block access to its Fable 5 and Mythos 5 models for foreign users over national security concerns.

OpenAI confirms GPT 5.6 Sol, Terra and Luna rollout after US govt review

OpenAI will make its GPT 5.6 AI models available to everyone this Thursday. The new lineup includes three models: GPT 5.6 Sol, Terra and Luna. The announcement comes weeks after the company introduced the models in a limited preview for a small group of trusted partners. At the time, OpenAI said it chose a limited release because the US government requested an early preview before the wider launch. The company has now confirmed that the public rollout will begin this week. According to OpenAI, Sol is the company's most powerful model yet, Terra is built for everyday use, and Luna focuses on speed and affordability. The company also claims that all three models come with its "most robust safeguards to date." OpenAI GPT 5.6 capabilities OpenAI describes GPT 5.6 Sol as its "strongest model" so far and also says it is the company's "most capable model yet for cybersecurity." OpenAI says Sol delivers better agentic capabilities across coding, biology and cybersecurity tasks. It is designed to help users identify and fix security vulnerabilities without crossing its internal safety boundaries. For users looking for a more balanced option, Terra offers performance similar to GPT 5.5 at half the price. On the other hand, Luna is aimed at users who need fast responses at a lower price. Also read: Meta breaks silence on Instagram child safety row, says offending ads and accounts have been removed When the GPT 5.6 series was first announced, OpenAI CEO Sam Altman explained why the company was taking a gradual approach to the rollout. He said the gradual release was a "quite reasonable" approach for increasingly capable AI systems, but added that this "isn't quite the process that we think is optimal." Altman also said OpenAI is working with the US government to create a more transparent and reliable process so future AI models can be released to users more quickly. Also read: Perplexity working on an AI coding tool to compete with Claude Code: Report The company also made it clear that it does not want government previews to become the normal way of launching AI models. OpenAI said, "It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them. We are taking this short-term step because we believe it is the strongest path to broader availability in the coming weeks, while we work with the Administration to develop the cyber Executive Order framework and a repeatable process for future model releases."