Subscribe

Reasoning models

Models and papers about reasoning, where a model works through a problem step by step before it answers. The Atlas has logged 343 launches and papers on it since 10 May 2023, 181 of them in 2026.

OpenAI, Google DeepMind and Anthropic have the most.

Key launches

The 60 most significant of 343, newest first. Every launch in the Atlas

  • AI-aided proof of finite-time blowup for the Navier-Stokes equations Anthropic · 21 November 2026
    NYU's Tristan Buckmaster and Anthropic's Levent Alpöge published an AI-aided proof of Navier-Stokes blowup. They say it closes a loophole in OpenAI's September solution.
  • Claude-assisted proof of the Riemann Hypothesis (Levent Alpöge) Anthropic · 20 November 2026
    Anthropic mathematician Levent Alpöge says he proved the Riemann Hypothesis and the Generalized Riemann Hypothesis with Claude, and posted a Lean formalization. It is not yet…
  • Time horizon plot update with judgment-heavy tasks METR · 20 November 2026
    METR updated its time horizon plot with longer and more open-ended tasks and found exponential growth continues, while models do worse on tasks that need judgment.
  • Inkling 2 Mini Thinking Machines · 20 November 2026
    Thinking Machines released Inkling 2 Mini, an open-weights looped mixture-of-experts model with 26.4B total parameters that runs on a laptop and scores 48.2% on DeepSWE v1.1.
  • Gemini 4 Neon Google DeepMind · 18 November 2026
    Google released Gemini 4 Neon, a mid-tier Gemini 4 model that scores 61 on the Artificial Analysis Intelligence Index and costs $3/$15 per million tokens.
  • GPT-6.2 Astra and Sol OpenAI · 17 November 2026
    OpenAI released GPT-6.2 Astra, its new frontier model, and GPT-6.2 Sol, an efficient model, with Astra API prices 20% below GPT-6 Astra.
  • A step toward defining the Standard Model on a lattice (GPT-6 Astra proof) OpenAI · 15 November 2026
    OpenAI says GPT-6 Astra took the lead in the detailed proof behind physicist Nathaniel Craig's candidate nonperturbative lattice definition of the Standard Model.
  • Gemini 4 Hydrogen Google DeepMind · 12 November 2026
    Google released Gemini 4 Hydrogen, a fast and cheaper Gemini 4 model for agentic and coding work, priced at $1.50 per million input tokens and $7.50 per million output tokens.
  • AlphaProof Code Google DeepMind · 10 November 2026
    Google DeepMind introduced AlphaProof Code, an agentic framework that formally verified the code of Signal's cryptographic protocol in Lean in four weeks.
  • Claude Fable 5.2 Anthropic · 9 November 2026
    Anthropic launched Claude Fable 5.2 and Claude Mythos 5.2, the same model, at $8 per million input tokens and $40 per million output tokens, 20% below version 5.1.
  • Claude Fable 5.2 and Claude Mythos 5.2 Anthropic · 9 November 2026
    Anthropic released Claude Fable 5.2 and Mythos 5.2, and says a team of 1,000 Fable 5.2 agents formalized Fermat's Last Theorem in 30 hours.
  • Proof of the Aaronson-Ambainis conjecture PSI · 6 November 2026
    Physical Superintelligence (PSI) says its AI system, running a non-public Anthropic model, proved the Aaronson-Ambainis conjecture, a quantum complexity question open since 2008.
  • GPT-6 Cyber OpenAI · 5 November 2026
    OpenAI released GPT-6 Cyber, a security model rated Critical under its Preparedness Framework, scoring 48.5% on ExploitGym, to verified Daybreak Red users first.
  • LongCat-2.5 Meituan · 2 November 2026
    Meituan released LongCat-2.5, a 1.6 trillion parameter native multimodal model with about 48 billion active parameters per token, built for long-horizon agentic coding.
  • Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights MIT CSAIL · 2 November 2026
    MIT CSAIL researchers report that GPT-6 Astra can read and write neural network weights, and its edit of 0.016% of Qwen3.5-0.8B's weights raised MATH-500 from 7.9% to 48.1%.
  • Gemini 4 Carbon Google DeepMind · 28 October 2026
    Google announced Gemini 4 Carbon, its strongest model, with top cybersecurity and long-horizon results, available first to a select group of cyber defenders before expanding to…
  • Refutation of the Exponential Time Hypothesis OpenAI · 27 October 2026
    OpenAI says its models found sub-exponential algorithms for a hard class of logic puzzles, which would refute the Exponential Time Hypothesis. Anthropic claimed its own…
  • GPT-6.1 Terra OpenAI · 26 October 2026
    OpenAI released GPT-6.1 Terra, a mid-tier model for computer use and writing, priced at $1 per million input and $5 per million output tokens, half the price of GPT-5.6 Terra.
  • Working with academic communities to build scientific understanding OpenAI · 26 October 2026
    OpenAI announced three more discoveries in math, computer science and physics, this time asking outside experts to act as shepherds who organize groups to understand the results…
  • Kimi K3.1 Pro Moonshot AI · 23 October 2026
    Moonshot AI released Kimi K3.1 Pro, a 2.8T-parameter MoE with 104B active parameters and 1M context, on its API today, with full weights promised by Nov 4, 2026.
  • PHP2Shell vulnerability discovery (Patch the Planet) OpenAI · 22 October 2026
    Vulnerabilities in core PHP, named PHP2Shell, were found by two parties, one a Calif researcher using GPT-5.6 Sol in OpenAI's Patch the Planet project, and disclosed after a…
  • MiniMax M3.1 Flash MiniMax · 19 October 2026
    MiniMax released M3.1 Flash, a Flash model for agentic work and coding that it says runs nearly three times faster and costs 70% less than its predecessor, at $0.15 input and…
  • Muse Spark Open Meta · 15 October 2026
    Meta released Muse Spark Open, an 816B-parameter open-weight mixture-of-experts model with 42B active parameters, which it calls the most capable US open-weight model to date.
  • Nemotron 3.5 Super VL NVIDIA · 15 October 2026
    NVIDIA released Nemotron 3.5 Super VL, an open 120B parameter multimodal mixture-of-experts model with 12B active parameters, built for long-running enterprise agents.
  • Muse Wonder and Muse Wonder Full Meta · 12 October 2026
    Meta launched Muse Wonder and Muse Wonder Full, its most capable models so far, aimed at personal and professional agents and available through paid Muse plans and the Meta Model…
  • Maia 300 Microsoft · 12 October 2026
    Microsoft announced Maia 300, its third-generation in-house AI chip built on TSMC N2P, to be the primary inference system for its products, with production rollout in 2027.
  • Claude Haiku 5.5 Anthropic · 7 October 2026
    Anthropic released Claude Haiku 5.5, its smallest model, with a 1M token context window at $0.10 per million input tokens and $0.50 output, about 75% cheaper than Haiku 4.5 on…
  • Sharing AI progress in mathematics (722 math manuscripts) OpenAI · 6 October 2026
    OpenAI published 722 math manuscripts in 372 result families, produced by an unreleased internal frontier model, in a public GitHub repository with Lean formalizations of many…
  • OpenAI math results release (722 manuscripts, 372 result groups) OpenAI · 6 October 2026
    OpenAI published 722 math manuscripts in 372 groups of related results on GitHub, produced by an unreleased internal model that attempted about 4,000 research problems.
  • 377 solved unsolved math problems OpenAI · 6 October 2026
    OpenAI published more than 700 papers describing solutions to 377 previously unsolved math problems, produced by an unreleased AI model.
  • Beam Reflection AI · 5 October 2026
    Reflection AI unveiled Beam, its first open-weight model, a 501B parameter mixture-of-experts model with 23B active parameters, aimed at coding and agent tasks.
  • Gemini 4 Argon Google DeepMind · 30 September 2026
    Gemini 4 Argon, a new frontier model with a 1M-token output limit, scores 77.9% on DeepSWE v1.1; released first to vetted cyber defenders via Fairwind.
  • GPT-Synopsys Synopsys / OpenAI · 30 September 2026
    Synopsys and OpenAI announced a multi-year partnership to build GPT-Synopsys, a specialized model that operates Synopsys EDA tools to run chip design workflows, with shared…
  • Claude Opus 5.5 Anthropic · 22 September 2026
    Claude Opus 5.5 matches Fable 5.1 on most work, costs 40% less than Opus 5, and scores 66.4% on Terminal-Bench 4.0.
  • GPT-6 Sol and GPT-6 Luna OpenAI · 22 September 2026
    GPT-6 Sol and Luna bring Astra-era training to cheaper tiers at half the price of their GPT-5.6 equivalents.
  • MiMo-V2.6-Pro and V2.6-Flash Xiaomi (MiMo) · 21 September 2026
    Omni-modal 1.02T (Pro) and 310B (Flash) MIT-licensed models; Artificial Analysis rated Pro 46, top of open-weight models at launch.
  • AI-generated Navier-Stokes finite-time blow-up proof (with Lean formalization) OpenAI · 8 September 2026
    OpenAI claims an internal agent system proved a Navier-Stokes finite-time singularity with smooth forcing, resolving the Millennium Prize problem, with a Lean proof.
  • Formalizing Fermat's Last Theorem Anthropic · 4 September 2026
    Claude agents wrote the first complete computer-checked Lean proof of Fermat's Last Theorem in about 11 days, with 13 million lines of Lean.
  • GPT-6 Astra OpenAI · 3 September 2026
    GPT-6 Astra, OpenAI's first model rated Critical for cybersecurity, claims 97.6% on FrontierMath Tier 4 and 99.9% on ARC-AGI-3.
  • Claude Fable 5.1 and Mythos 5.1 Anthropic · 1 September 2026
    Claude Fable 5.1 lifts Terminal-Bench 4.0 from 42.0% to 55.8% and cuts cache-read price 75%; Mythos 5.1 is the same model with fewer safeguards.
  • Hy4 preview Tencent · 28 August 2026
    770B-total, 49B-active open MoE with 1M-token context, Gated DeepSeek Sparse Attention and hyper-connections, under Apache-2.0.
  • Qwen3.8-2.4T-A95B (open weights) Alibaba (Qwen) · 12 August 2026
    Open weights of a 2.4T-parameter, 95B-active MoE: Terminal Bench 2.1 86.6 and SWE-bench Pro 67.7, under a Qwen3.8-Max license.
  • Claude raises the lower bound on Riemann zeta zeros on the critical line Anthropic · 10 August 2026
    An unreleased Claude raised the proven fraction of Riemann zeta zeros on the critical line from 41.6% to 67.2% while attempting the Riemann hypothesis.
  • Qwen3.8-Max Alibaba (Qwen) · 2 August 2026
    Alibaba's largest model has 2.4T parameters (95B active) and 1M context, and is positioned for coding and 'cowork'. Open weights followed the next week.
  • Ten advances in mathematics and theoretical computer science OpenAI · 1 August 2026
    An internal Astra found ten results on decade-old open problems, including non-sofic groups, each with a Lean certificate.
  • Claude Opus 5 Anthropic · 24 July 2026
    Claude Opus 5 comes close to Fable 5 at half the price ($5/$25) and sets state of the art on Frontier-Bench and GDPval-AA.
  • Kimi K3 Moonshot AI · 16 July 2026
    2.8T-parameter open-weights MoE (104B active, 1M context) on Kimi Delta Attention and Attention Residuals, reported third on Artificial Analysis behind two closed models at launch.
  • GPT-5.6 (Sol, Terra, Luna) OpenAI · 9 July 2026
    GPT-5.6 splits into Sol, Terra and Luna tiers, with Sol scoring 53.6 on Agents' Last Exam and 92.2% on BrowseComp at lower token cost.
  • Hy3 Tencent · 6 July 2026
    Completed Hy3: same 295B/21B MoE, now Apache-2.0, API input at 1 yuan per million tokens; scored 2.67 vs GLM-5.1's 2.51 in a 270-expert blind test.
  • Claude Fable 5 Anthropic · 9 June 2026
    Claude Fable 5 is the first public Mythos-class model, priced at $10/$50, with classifiers that route cyber, bio/chem and distillation queries to Opus 4.8.
  • Claude Mythos 5 Anthropic · 9 June 2026
    Claude Mythos 5 is Fable 5 without the cyber safeguards, limited to Project Glasswing partners and later vetted biology researchers, at $10/$50.
  • Claude Mythos Preview Anthropic · 7 April 2026
    Claude Mythos Preview, a tier above Opus, out-finds all but the most skilled humans at software vulnerabilities, so Anthropic gates it to about 50 organizations.
  • Gemini 3 Pro Google DeepMind · 18 November 2025
    Gemini 3 Pro scored 1501 Elo on LMArena, 37.5% on Humanity's Last Exam, 91.9% on GPQA Diamond and 76.2% on SWE-bench Verified, and shipped in Search on day one.
  • Gemini Deep Think at IMO 2025 (gold-medal score) Google DeepMind · 21 July 2025
    Advanced Gemini Deep Think scored 35/42 at IMO 2025, a gold-medal score, working end to end in natural language within 4.5 hours; IMO-graded.
  • DeepSeek-R1: Incentivizing Reasoning via RL (arXiv) DeepSeek · 22 January 2025
    Shows reasoning can emerge from pure RL on verifiable rewards (R1-Zero), then builds R1 with a small cold-start and distills it to small models.
  • DeepSeek-R1 and R1-Zero DeepSeek · 20 January 2025
    Open-weights 671B MoE reasoning model claimed at OpenAI o1 level; R1-Zero showed reasoning emerging from pure RL with rule-based rewards.
  • o3 (announcement) OpenAI · 20 December 2024
    o3 scores 75.7% (high-efficiency) and 87.5% (low-efficiency) on ARC-AGI-1 semi-private tasks; announced on day 12 of OpenAI's December event, not released.
  • o1-preview and o1-mini OpenAI · 12 September 2024
    First model trained with large-scale reinforcement learning to think in a long hidden chain of thought before answering.
  • Gemini 1.5 Pro Google DeepMind · 15 February 2024
    Mixture-of-experts Gemini 1.5 Pro matched 1.0 Ultra with less compute and offered a 1M-token context (10M tested in research).
  • Gemini 1.0 (Ultra, Pro, Nano) Google DeepMind · 6 December 2023
    First Gemini family, natively multimodal, with Ultra, Pro and Nano models; Ultra reported 90.0% on MMLU with CoT@32, beating GPT-4's reported score.

Most active labs