Google Releases Gemini 3.8 Flash AI Model, Third Launch in Six Weeks With Cybersecurity Focus
22 Sources
[1]
Google releases Gemini 3.8 Flash, its third Flash model in six weeks
Google hasn't released a frontier-level Gemini Pro AI model since early 2026, but it sure loves rolling out new Gemini Flash variants. Today, Google is announcing its third Flash model release in just six weeks, making it more likely that we'll never see the promised Gemini 3.5 Pro. But no matter,
[2]
Google says its new Gemini 3.8 Flash model 'works harder' but might cost more
Google launched Gemini 3.8 Flash, arriving just a few weeks after its predecessor. The company claims the new model "works harder" than Gemini 3.7 Flash by performing more reasoning steps on complex tasks and "calling tools iteratively." It has the same introductory pricing as 3.7 Flash, $0.75 per
[3]
With Gemini 3.8 Flash, Google reminds everyone it's still in the race
Google on Wednesday announced the release of Gemini 3.8 Flash in an attempt to reclaim its reputation as a top-tier model maker in the process. The company's commitment to the frontier model race has been in doubt since at least early August when Google DeepMind CEO Demis Hassabis stepped down to
[4]
Google releases Gemini 3.8 Flash and a cybersecurity variant limited to governments
Gemini 3.8 Flash arrives three weeks after 3.7 with a cybersecurity variant limited to trusted testers and unnamed governments, while the AI Act attaches documentation, copyright and training data obligations to every general purpose model placed on the EU market Google has released Gemini 3.8
[5]
Gemini 3.8 Flash is out with a new focus on cybersecurity
Gemini 3.8 Flash Cyber brings specialized vulnerability detection and automated patching capabilities, but access is limited to Google's Fairwind Program. Google's previous smartest AI model, Gemini 3.6 Flash, was released in late July. The tech giant wasted no time iterating on it, with Gemini
[6]
Right on Cue, Google Is Reportedly on the Verge of Shipping a Coding-First AI Model
According to the popular narrative right now, Google is behind when it comes to AI models. In the popular imagination, it finished at the top of the pile last year, and then got steamrolled by "agents" when platforms like OpenClaw made token-guzzling agentic coding into the AI trend of 2026. So if
[7]
Gemini 3.8 Flash rolling out three weeks after last release
After the last model release three weeks ago, Google today is rolling out Gemini 3.8 Flash. Our most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows. 3.8 Flash beats its predecessor in various benchmarks: Like the
[8]
Gemini 3.8 Flash has one major advantage over GPT-6 Astra and Claude Fable 5.1
Google's workhorse AI just outmatched OpenAI on video capabilities This month has been a whirlwind for AI, with major new models arriving within days of one another. Google just released Gemini 3.8 Flash, and at first glance it has a tough job ahead of it. GPT-6 Astra is OpenAI's most capable
[9]
Gemini 3.8 Flash could land any day now, and it could put the vibe back into vibe coding
Google has internally scrapped plans to launch Gemini 3.5 Pro, and could only release the next Pro model with Gemini 4.0. With such a wide range of applications, Gemini feels more like a general-purpose technology or GPT (no, not that GPT) and less like a specialized frontier model. But it has
[10]
'I also want to feel the frontier' -- Gemini users are starting to think that Gemini Pro 4 won't ever see the light of day thanks to the release of ChatGPT 6 Astra and Claude Fable 5.1
Being a devoted Gemini user seems to require a peculiar combination of loyalty, patience, and an unusually high tolerance for the word Flash. That patience is being tested again. Anthropic released Claude Fable 5.1 last week, bringing significant improvements to its model lineup. OpenAI followed
[11]
Google's Gemini 3.8 Flash is built for agents, while its Cyber twin hunts vulnerabilities
Google keeps cranking out Flash models: the company on Wednesday announced two versions of a new 3.8 Flash. The variants include a standard Flash, a "workhorse" model for agentic tasks, software development, and multi-step reasoning, and Flash Cyber optimized for vulnerability detection and
[12]
Google shipped four Gemini Flash models in 106 days. Yet its Gemini 3.5 Pro is still AWOL. | Fortune
Google expected Gemini 3.5 Pro to ship in June. It was still undergoing testing in July, and Google's website still lists it as "coming soon." Internal candidates were discarded because they did not improve enough over Flash, the Wall Street Journal reported. Google, however, has begun pointing to
[13]
Google launches Gemini 3.8 Flash and Gemini 3.8 Flash Cyber
Gemini 3.8 Flash targets coding and agentic tasks, while the Cyber variant focuses on vulnerability detection and patching Google $GOOGL introduced two AI models on Wednesday: Gemini 3.8 Flash, a general-purpose model aimed at software engineering and autonomous tasks, and Gemini 3.8 Flash Cyber,
[14]
I Tried Gemini 3.8 Flash, Google's Latest AI Upgrade, and It (Mostly) Lives Up to the Hype
There are noticeable improvements over the earlier Flash 3.6 model. Google has rolled out two new AI models: Gemini 3.8 Flash (the "intelligent workhorse model" for everyday prompting), and Gemini 3.8 Flash Cyber (available to "trusted defenders" for cybersecurity protection). While we only got
[15]
Google launches two Gemini 3.8 models with cutting-edge reasoning capabilities
Google launches two Gemini 3.8 models with cutting-edge reasoning capabilities Three weeks after its last large language model release, Google LLC today launched Gemini 3.8 Flash and Gemini 3.8 Flash Cyber. The two LLMs are based on the same technical foundation. The primary difference is that
[16]
Google Gemini 3.8 Flash coding AI model expected this week
Gemini 3.8 Flash, known internally as "Skimaki," is expected to release Wednesday, according to the Wall Street Journal Google $GOOGL's AI research unit plans to ship a new model boasting sharply improved coding abilities as soon as Wednesday, the Wall Street Journal reported. The model, Gemini
[17]
Gemini 3.8 Flash Outperforms GPT 5.6 Sol in Coding Tests
Google's latest AI model, Gemini 3.8 Flash, introduces significant advancements in reasoning, coding and multimodal functionality. Released in two variants, Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, this model addresses diverse applications, including front-end design automation and real-time
[18]
Google Gemini 3.8 Flash Costs $0.75 per Million Input Tokens
Google's latest release, Gemini 3.8 Flash, marks a significant addition to the AI landscape, offering specialized capabilities tailored to specific industries. As highlighted by Matthew Berman, the model excels in targeted benchmarks such as the Harvey Legal Benchmark, where it achieves a leading
[19]
Google rolls out Gemini 3.8 Flash and 3.8 Flash Cyber for coding, reasoning and cyber defense
Google has introduced Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, its latest models focused on software engineering, agentic tasks, multi-step reasoning, vulnerability detection and automated patching. Gemini 3.8 Flash is designed for general-purpose workloads, while Gemini 3.8 Flash Cyber focuses
[20]
New Google Gemini 3.8 Flash Matches Opus 5 in Deep Sweep 1.1 Benchmark
Google's recent release of Gemini 3.8 Flash has introduced a model that balances advanced performance with cost efficiency, catering to both technical and enterprise needs. As highlighted by Prompt Engineering, this model excels in specific benchmarks like Deep Sweep 1.1, demonstrating its ability
[21]
Google launches Gemini 3.8 Flash Cyber for government security By Investing.com
Investing.com -- Google announced a pair of upgraded artificial-intelligence models on Wednesday, sharpening its focus on government and enterprise cybersecurity in a bid to insulate its AI revenues from industry-wide price pressures The tech giant released Gemini 3.8 Flash, an upgraded version of
[22]
Google introduces Gemini 3.8 AI models for agentic workflows and cybersecurity: Here is how to access them
Google describes Gemini 3.8 Flash as its "most intelligent workhorse model." Google has introduced Gemini 3.8 AI models for agentic workflows and cybersecurity. The company has launched two versions: Gemini 3.8 Flash and Gemini 3.8 Flash Cyber. Google says the new models offer stronger performance
Share
Copy Link
Google launched Gemini 3.8 Flash, its third Flash AI model in six weeks, featuring enhanced software engineering capabilities and a specialized cybersecurity variant. The model achieves 73.7% on the DeepSWE benchmark while maintaining competitive pricing at $0.75 per million input tokens.
Google Accelerates AI Model Releases With Gemini 3.8 Flash
Google has launched
1
1
, marking its third Flash AI model release in just six weeks. The rapid deployment comes as the company has not released a frontier-level Gemini Pro model since early 2026, raising questions about the future of the promised Gemini 3.5 Pro. The new release arrives in two variants: the standard Gemini 3.8 Flash, which Google describes as a "workhorse" model for agentic tasks and software engineering, and Gemini 3.8 Flash Cyber, tuned specifically for vulnerability detection and mitigation in cybersecurity applications2
.
Source: Android Authority
The accelerated release cadence reflects Google's strategy to maintain competitiveness through rapid iteration rather than waiting for major flagship updates. For developers and enterprise users, this approach delivers incremental improvements while keeping the company visible in an increasingly crowded AI model marketplace where rivals have been aggressively cutting prices to retain wary business customers
3
.Benchmark Performance and Software Engineering Capabilities
Gemini 3.8 Flash demonstrates significant improvements in software engineering tasks, achieving 73.7% on the
1
1
, which measures a model's ability to autonomously solve complex engineering problems. This score positions it at the top of the DeepSWE v1.1 leaderboard, outperforming larger and more expensive frontier models including Claude Sonnet 5, GPT-5.6 Sol, and GPT-5.6 Terra, while falling just below Claude Opus 5's 74% score5
.The model also excels across multiple specialized benchmarks. On the Vals Finance Agent V2 benchmark for financial analysis, Harvey's Legal Agent benchmark for complex legal workflows, and CharXiv Reasoning for information synthesis, Gemini 3.8 Flash outperformed its competitors
5
. According to2
2
, the model scores 59 on the Intelligence Index, putting it level with GPT-5.6 Sol and Grok 4.6, while remaining "the cheapest we've measured at this level of intelligence."Google explains that the performance gains come from the model "working harder" by executing extra reasoning steps and calling tools iteratively on complex tasks
3
. This increased diligence means Gemini 3.8 Flash may use more tokens to maximize performance, especially at higher effort levels, potentially increasing costs despite unchanged per-token pricing2
.Pricing Strategy and Market Positioning
Google is offering
1
1
to Gemini 3.8 Flash at2
2
of $0.75 per million input tokens and $3.75 per million output tokens through the end of the year. After this promotional period, pricing will double to $1.50 per million input tokens and $7.50 per million output tokens1
. Despite the unchanged per-token pricing, actual costs could increase by approximately 40% compared to Gemini 3.7 Flash due to a 30% increase in output tokens per task and more turns on agentic evaluations2
.At $0.58 per Intelligence Index task, the model remains significantly more affordable than competitors. Anthropic's Fable 5.1 costs approximately $3.76 per task—roughly 6 times more expensive
3
. This pricing advantage positions Google competitively against rivals who have recently upgraded their offerings, including Anthropic's Fable 5, which received an enhancement earlier this week promising stronger performance at lower prices through reduced costs for cached data2
.Developers seeking to minimize
2
2
can continue using Gemini 3.7 Flash, giving them flexibility based on their specific use cases and budget constraints.Gemini 3.8 Flash Cyber and the Fairwind Program
4
4
represents Google's specialized offering for cybersecurity applications, replacing the 3.5 version. Access to this variant is restricted through the newly launched5
5
, which is limited to1
1
, critical infrastructure operators, and software maintainers. The program's 650 members include1
1
,1
1
, CrowdStrike, and the Center for Internet Security2
.
Source: SiliconANGLE
The model achieved 86.2% on the industry-standard benchmark for finding vulnerabilities in C/C++, surpassing its predecessor Gemini 3.5 Flash Cyber and larger models like Mythos 5 and GPT-5.5-Cyber
5
. In real-world5
5
spanning over 20 programming languages, it reached a 71% success rate, significantly higher than Gemini 3.7 Flash's 58.9% and Gemini 3.5 Flash Cyber's 46.6%5
.Google's internal testing revealed substantial improvements in
1
1
. The Chrome security team reported a 2.6x increase in correct patches compared to the best commercial models that are much larger5
. The Cloud team found that 3.8 Flash Cyber identified a critical vulnerability in just two hours1
. The model also works with Google's2
2
agent to autonomously find and fix vulnerabilities, protecting critical infrastructure, public services, and national security.Related Stories
Regulatory Compliance and European Market Challenges
Each Gemini 3.8 Flash release creates compliance obligations under the
4
4
. Article 53 requires technical documentation, a copyright policy, and a public training data summary for every general purpose AI model placed on the EU market4
. Google voluntarily joined the general purpose AI Code of Practice on July 30, 2025, a week after Meta declined4
.For models classified as systemic risk—those trained above 10 to the 25th floating point operations—Article 52 mandates notification to the European Commission within two weeks
4
. This window is shorter than the three-week gap between Gemini 3.7 Flash and 3.8 Flash releases. Whether any Flash models cross this threshold remains undisclosed, as Google has not publicly shared training compute details, and the determination depends on training compute rather than benchmark results4
.The rapid release cadence creates a documentation burden, as each model launch triggers separate European obligations for a product line that was initially unavailable in Europe at launch
4
. The restricted access to Gemini 3.8 Flash Cyber raises additional questions, as Google has not disclosed which governments receive access to the specialized cybersecurity variant4
.Availability and Access Options
Gemini 3.8 Flash is available immediately across Google's ecosystem through multiple channels. Developers can access it via Google Antigravity, the Gemini API in Google AI Studio and Android Studio, and interface design service Stitch
3
. Enterprise customers can use Gemini Enterprise, while1
1
can access the model through the Gemini app, AI Mode in Google Search, and Gemini in Google Sheets3
. Users wanting to experiment with the model for free can visit AI Studio1
.
Source: Android Authority
The model's improvements in agentic workflows have drawn positive early reactions. Aigora.ai CEO John Ennis compared it favorably to Anthropic's models, stating it offers "Opus 5 coding quality but at a fraction of the cost and super fast," adding that "This is going to be so awesome for things like making remotion videos"
2
. However, computer use remains a challenge—while Gemini 3.8 Flash shows improvement over 3.7 Flash in the OSWorld-2.0 test of agentic computer use, it still lags significantly behind market leader Claude Opus1
.References
Summarized by
Navi
[5]
Related Stories
Google launches Gemini 3 Flash as default AI model, delivering speed with Pro-grade reasoning
17 Dec 2025•Technology
Google Unveils Gemini 2.5 Flash: A Faster, More Efficient AI Model
08 Apr 2025•Technology
Google launches Gemini 3.1 Flash Lite with 2.5x speed boost and adjustable thinking levels
03 Mar 2026•Technology
Recent Highlights
1
AI Leaders Warn UN Security Council of Existential Risks, Call for Global AI Governance
Policy and Regulation
2
AI Hallucination Nearly Sparked US Military Confrontation With Chinese Ship in Middle East
Technology
3
OpenAI AI Agent Breaches Australian Government Healthcare Portal in Historic Hack
Technology
Recent Highlights
Today's Top Stories
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.
