DeepSeek releases open-source AI models that rival GPT-5 and Gemini at fraction of the cost
13 Sources
[1]
DeepSeek Releases New Reasoning Models to Take On ChatGPT and Gemini
Imad is a senior reporter covering Google and internet culture. Hailing from Texas, Imad started his journalism career in 2013 and has amassed bylines with The New York Times, The Washington Post, ESPN, Tom's Guide and Wired, among others. DeepSeek, the China-based AI company, has released two new
[2]
Is DeepSeek's new model the latest blow to proprietary AI?
Chinese AI firm DeepSeek has made yet another splash with the release of V3.2, the latest iteration in its V3 model series. Launched Monday, the model, which builds on an experimental V3.2 version announced in October, comes in two versions: "Thinking," and a more powerful "Speciale." DeepSeek
[3]
DeepSeek just gave away an AI model that rivals GPT-5 - and it could change everything
Their release challenges U.S. tech dominance and reignites global AI competition Chinese startup DeepSeek has casually lobbed two new enormous AI models into the already smoldering international arms race: DeepSeek-V3.2, a model built for everyday reasoning, and DeepSeek-V3.2-Speciale, a
[4]
DeepSeek just dropped two insanely powerful AI models that rival GPT-5 and they're totally free
Chinese artificial intelligence startup DeepSeek released two powerful new AI models on Sunday that the company claims match or exceed the capabilities of OpenAI's GPT-5 and Google's Gemini-3.0-Pro -- a development that could reshape the competitive landscape between American tech giants and their
[5]
DeepSeek Releases New Reasoning Models to Match GPT-5, Rival Gemini 3 Pro | AIM
Both models and the accompanying technical report have been released as open source on Hugging Face. Chinese AI lab DeepSeek has launched two new reasoning-first AI models, DeepSeek-V3.2 and DeepSeek-V3.2-Speciale, expanding its suite of systems for agents, tool-use and complex inference. Both
[6]
How DeepSeek 3.2 Uses Specialists & Improved Memory to Outthink Gemini 3.0 Pro
What if we told you that an open source AI just outperformed one of the most advanced proprietary models on the market? Yes, you read that right -- DeepSeek 3.2 has officially outclassed Gemini 3.0 Pro, a feat many thought impossible just a few years ago. For decades, open source AI has been seen
[7]
DeepSeek's New AI Model Achieves Gold-Level Results, Rivals GPT-5 and Gemini 3 Pro
The DeepSeek V3.2-Speciale model delivers gold-level results in IMO, CMO, ICPC, and IOI 2025. China's frontier AI lab, DeepSeek has released its new reasoning AI models that rival the performance of OpenAI's GPT-5 High and Google's Gemini 3 Pro. There are two AI models including DeepSeek-V3.2 and
[8]
DeepSeek 'Integrates Thinking' Into New AI Models | PYMNTS.com
By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. "DeepSeek-V3.2 is our first model to integrate thinking directly into tool-use,
[9]
DeepSeek launches pair of AI models to challenge OpenAI, Google
China-based DeepSeek (DEEPSEEK) has launched a pair of new artificial intelligence models, DeepSeek-V3.2 and DeepSeek-V3.2-Speciale, which are open-sourced and topped or matched the results of OpenAI's (OPENAI) GPT-5 and Google's (GOOG)(GOOGL) Gemini 3 Pro in some benchmarks. DeepSeek-V3.2 is
[10]
DeepSeek 3.2 Challenges GPT-5 While Slashing AI Spend : IMO Gold to 128k Context
What if innovative AI didn't have to come with a sky-high price tag? Imagine an open source model that not only rivals proprietary giants like GPT-5 but also delivers gold medal-level performance in global competitions, all while being 30 times more affordable. Enter DeepSeek 3.2, a new leap in
[11]
DeepSeek 3.2 AI Outperforms GPT-5 & Gemini 3 Thanks to a New Bold Training Method
What does it take to outshine giants in the fiercely competitive world of artificial intelligence? For years, proprietary systems like GPT-5 and Gemini Pro have dominated the landscape, setting benchmarks others could only chase. Yet, against all odds, DeepSeek has done it again. With the release
[12]
New Deepseek 3.2 AI Open Model Outthinks ChatGPT 5 in Tough Reasoning Tests
What if the next leap in artificial intelligence wasn't locked behind corporate walls, but instead, freely available to everyone? That's the bold promise of Deepseek 3.2, the latest evolution in open source AI. With its jaw-dropping gold medal wins at both the International Math Olympiad (IMO) and
[13]
DeepSeek introduces new AI models that rival GPT-5 and Gemini 3.0: Here's what it offers
The launch marks DeepSeek's closest move yet toward AGI, after earlier versions shook global tech stocks. After Google and OpenAI introduced Gemini 3.0 and GPT-5, China's AI startup DeepSeek has introduced two new large language models, DeepSeek-V3.2 and DeepSeek-V3.2 Speciale, in order to keep
Share
Copy Link
Chinese AI company DeepSeek launched two new reasoning models, V3.2 and V3.2-Speciale, claiming performance on par with OpenAI's GPT-5 and Google's Gemini 3.0 Pro. The open-source models cost $0.028 per million tokens compared to Gemini 3's $4.00, and achieved gold-level performance in elite international math and coding competitions. The release challenges American tech dominance while making frontier AI capabilities freely accessible.
DeepSeek Launches New Reasoning Models to Challenge Tech Giants
The Chinese AI company DeepSeek has released two new reasoning models, DeepSeek V3.2 and DeepSeek V3.2-Speciale, positioning them as direct competitors to OpenAI's GPT-5 and Google's Gemini 3.0 Pro
1
. The Hangzhou-based startup announced Monday that V3.2 is now available on its app, web interface, and API, while the more powerful V3.2-Speciale variant can only be accessed through a temporary API endpoint until December 15, 20255
. Both AI models have been released under an open-source MIT license, with weights accessible via Hugging Face, continuing DeepSeek's pattern of making frontier AI systems freely available2
.
Source: Seeking Alpha
According to company benchmarks, DeepSeek V3.2 surpasses GPT-5 in reasoning performance and matches Gemini 3.0 Pro on key metrics
1
. The V3.2-Speciale variant achieved even more striking results, earning gold-medal status at four elite international competitions: the 2025 International Mathematical Olympiad, the International Olympiad in Informatics, the ICPC World Finals, and the China Mathematical Olympiad4
. These achievements position the models among the most capable systems currently available, directly challenging American tech dominance in AI development.Open-Source AI Models Slash Computational Costs Dramatically
What sets these new reasoning models apart is their cost efficiency. Accessing Gemini 3 through the API costs up to $4.00 per million tokens, while DeepSeek V3.2-Speciale runs at just $0.028 per million tokens
2
. Processing 128,000 tokens—roughly equivalent to a 300-page book—now costs approximately $0.70 per million tokens for decoding, compared to $2.40 for the previous V3.1-Terminus model, representing a 70% reduction in inference costs4
. This dramatic cost reduction threatens to undermine the substantial investments that proprietary labs like OpenAI, Google, and Anthropic pour into their models.The 685-billion-parameter models support context windows of 128,000 tokens, making them suitable for analyzing lengthy documents, codebases, and research papers
4
. Despite the massive scale—DeepSeek noted the model can only run on giant server stacks with millions of dollars in hardware—the open source nature means any company can access the model weights for free1
. This accessibility enables solo developers and student teams to work with systems that just months ago would have required a lab and substantial cloud budget3
.DeepSeek Sparse Attention Breakthrough Drives Performance Gains
At the core of DeepSeek's efficiency lies DeepSeek Sparse Attention (DSA), an architectural innovation that dramatically reduces the computational burden of running AI models on long documents and complex tasks
4
. Traditional attention mechanisms scale poorly as input length increases—processing a document twice as long typically requires four times the computation. DSA breaks this constraint using a "lightning indexer" that identifies only the most relevant portions of context for each query, essentially skimming rather than reading every word3
.
Source: VentureBeat
The company deployed this approach in two phases: an initial high-level scan of tokens in training data to identify the small subset most relevant to a particular query, then drilling into that subset with full computational power
2
. According to DeepSeek's technical report, DSA reduces inference costs by roughly half compared to previous models when processing long sequences, slashing costs for long documents by up to 70%3
. Independent evaluations on long-context benchmarks show V3.2 performing on par with or better than its predecessor despite incorporating the sparse attention mechanism4
.Rivals GPT-5 and Gemini 3.0 Pro on Elite Benchmarks
DeepSeek's claims of parity with leading American AI systems rest on extensive testing across mathematics, coding, and reasoning tasks. On AIME 2025, a prestigious American mathematics competition, DeepSeek V3.2-Speciale achieved a 96.0% pass rate, compared to 94.6% for GPT-5-High and 95.0% for Gemini 3.0 Pro
4
. On the Harvard-MIT Mathematics Tournament, the Speciale variant scored 99.2%, surpassing Gemini's 97.5%. The standard V3.2 model scored 93.1% on AIME and 92.5% on HMMT—marginally below frontier AI systems but achieved with substantially fewer computational resources4
.The competition results proved especially striking. DeepSeek V3.2-Speciale scored 35 out of 42 points on the 2025 International Mathematical Olympiad, earning gold-medal status, and scored 492 out of 600 points at the International Olympiad in Informatics—also gold, ranking 10th overall
4
. These results came without internet access or external tools during testing. On coding benchmarks, DeepSeek V3.2 resolved 73.1% of real-world software bugs on SWE-Verified, competitive with GPT-5-High at 74.9%, while on Terminal Bench 2.0, measuring complex coding workflows, DeepSeek scored 46.4%—well above GPT-5-High's 35.2%4
.Related Stories
Thinking in Tool-Use Transforms Agent Capabilities
Beyond raw reasoning, DeepSeek V3.2 introduces "thinking in tool-use"—the ability to reason through problems while simultaneously executing code, searching the web, and manipulating files
4
. Previous AI models faced a frustrating limitation: each time they called an external tool, they lost their train of thought. DeepSeek fixed this by preserving memory across tools, training the model using over 85,000 complex synthetic instructions to work with tools like real web browsers and coding environments3
.The models represent an expansion of DeepSeek's agent training approach, supported by a new synthetic dataset spanning more than 1,800 environments
5
. This level of real-world task preparation enables capabilities that most current chatbots simply aren't built for—planning multi-day vacations under strict budgets with interdependent constraints while testing code snippets and checking exchange rates simultaneously3
. The company stated that V3.2 is its first model to integrate thinking directly into tool use, allowing structured reasoning to operate both within and alongside external tools5
.Global AI Competition Intensifies Amid Geopolitical Tensions
DeepSeek's release carries profound implications for American technology leadership and the broader AI competition. The company has demonstrated it can produce frontier AI systems despite U.S. export controls that restrict China's access to advanced Nvidia chips
4
. The company first gained worldwide attention earlier this year when it released a reasoning-capable model online for free, immediately turning the narrative about reasoning models on its head by showing it's possible to run smarter, more capable AI models at a fraction of the cost1
. This spooked Wall Street, with some wondering if OpenAI, Google, and Anthropic weren't innovating fast enough.
Source: TechRadar
Since then, ChatGPT, Gemini, and Claude have all released reasoning-level models for free, with higher-level "thinking" models available for paid subscribers
1
. However, openness doesn't mean transparency. German regulators have tried to block DeepSeek over data transfer concerns, Italy banned the app earlier this year, and U.S. lawmakers want it off government devices entirely3
. The geopolitical context looms large, raising questions about what American firms offer that justifies their markup if DeepSeek's models deliver frontier performance at dramatically lower costs while remaining freely accessible under the MIT license.References
Summarized by
Navi
[4]
Related Stories
Recent Highlights
1
OpenAI's AI Math Breakthrough on Millennium Problem Sparks Fierce Attribution Battle
Science and Research
2
Trump Dismisses AI Regulation Calls, Targets Anthropic CEO Amid Growing Safety Concerns
Policy and Regulation
3
Nvidia CEO Jensen Huang Declares AGI Has Arrived With OpenAI's GPT-6 Astra Release
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.
