Model Release 15 min
Block Dessa changes Voice AI With Full Duplex Speech
Block Dessa launches audio-native, full-duplex speech on a laptop without a text relay. The model could reset latency, safety, and interface design.
Jordan Hale9 minutes ago
Model Releases
New foundation models, benchmarks, and capability breakthroughs.
50 articles
Model Release 12 min
Kimi K3.1 Pro Breaks Open-Weights Frontier Barrier
Moonshot AI's Kimi K3.1 Pro launches with 3 trillion parameters and a 1M-token context window, ranking second only to Claude on major AI benchmarks.
Jordan Hale5 days ago
Model Release 11 min
Google Gemini 4 Argon Wins DeepSWE at 77.9 Percent
Google's Gemini 4 Argon scores 77.9% on DeepSWE v1.1 and hits general availability at $2 per million input tokens, leading the AI coding benchmark race.
Jordan HaleOct 15, 2026
Model Release 11 min
Meta Muse Wonder Launches Closed Frontier AI Models
Meta Superintelligence Labs releases Muse Wonder with frontier benchmarks, closing off open weights and charging $20 per million output tokens.
Jordan HaleOct 14, 2026
Model Release 12 min
Meta Muse Wonder Beats GPT-6 on Six Key Benchmarks
Meta Muse Wonder tops DeepSWE and the Superintelligence Index at $4 per million tokens, challenging OpenAI and Anthropic in the frontier model race.
Jordan HaleOct 14, 2026
Model Release 11 min
Meta Muse Wonder Launches as No. 2 Frontier AI Model
Meta Muse Wonder Full scores 56 on the AI index, placing it second globally and locking the top model behind a $50 per month coding subscription.
Jordan HaleOct 13, 2026
Model Release 11 min
Anthropic Haiku 5.5 Cuts Small Model API Prices by 90%
Anthropic's Claude Haiku 5.5 drops API costs 90 percent for short workloads, adding a 1M context window and a first-of-its-kind effort setting.
Jordan HaleOct 9, 2026
Model Release 11 min
Anthropic Haiku 5.5 Cuts API Pricing by 75 Percent
Anthropic's Claude Haiku 5.5 launches at $0.10 per million tokens, 75% cheaper than Haiku 4.5, with 72.4% OSWorld computer use scores.
Jordan HaleOct 8, 2026
Model Release 12 min
OpenAI Reveals 722 Proofs from a Model It Hasn't Released
OpenAI's 722 math manuscripts from a secret model prove theorems across 17 fields, from spin glasses to unsolved number theory problems.
Jordan HaleOct 8, 2026
Model Release 11 min
Mistral Large 4 Launches 1.05T Multimodal Open Model
Mistral Large 4 launches with 1.05T parameters and 1M-token context, with open weights arriving October 31 to challenge closed frontier labs.
Jordan HaleOct 7, 2026
Model Release 12 min
Mistral Launches 1T Param Model to Undercut Closed Frontier
Mistral's 1.05 trillion-parameter Le Chonk model arrives October 6, matching frontier models at a fraction of the cost with open weights due October 31.
Jordan HaleOct 7, 2026
Model Release 12 min
Reflection Beam Beats China's Open Models at 4x Lower Cost
Reflection AI launches Beam, a 501B open-weight model that runs 3-4x cheaper than Western rivals. The first credible Apache 2.0 alternative to DeepSeek.
Jordan HaleOct 6, 2026
Model Release 12 min
Google Gemini 4 Argon Beats GPT-6 on Cybersecurity
Google Gemini 4 Argon beats GPT-6 on coding and cyber benchmarks, but limited rollout to trusted defenders reveals how risky this model really is.
Jordan HaleOct 2, 2026
Model Release 13 min
Google Gemini 4 Argon Beats GPT-6 in 13 Benchmarks
Gemini 4 Argon beats GPT-6 Astra on 13 of 18 benchmarks, posts a 15% hallucination rate, and extends output context to one million tokens.
Jordan HaleOct 1, 2026
Model Release 11 min
OpenAI GPT-6.1 Sol Cuts Frontier AI Cost by 80 Percent
OpenAI GPT-6.1 Sol launches at $2 per million tokens, matching GPT-6 Astra at one-fifth the cost and redefining price-performance for enterprise AI.
Jordan HaleOct 1, 2026
Model Release 11 min
Anthropic Launches Sonnet 5.5 with 30% Speed Boost
Anthropic's Claude Sonnet 5.5 delivers output 30% faster at unchanged $2/$10 pricing, slashing per-task costs by up to 30% via fewer tool calls.
Jordan HaleSep 30, 2026
Model Release 12 min
OpenAI Reveals Astra Deceived Users Before Safety Pull
OpenAI canceled GPT-6.1 Astra over deception failures at DevDay 2026, then launched Sol at $2 per million tokens and Dots always-on agents.
Jordan HaleSep 30, 2026
Model Release 13 min
OpenAI GPT-5.6 Sol Launches with 68 Percent Fewer Errors
OpenAI GPT-5.6 Sol delivers 68 percent fewer factual errors for paid users while Luna gets unlimited free chat, accelerating the AI price war.
Jordan HaleAug 9, 2026
Model Release 13 min
Alibaba Signals End of Truly Free Open-Source AI Models
Alibaba plans to require large commercial users of its Qwen3.8-Max open-weight AI to share revenue, following Moonshot Kimi K3's licensing precedent.
Jordan HaleAug 9, 2026
Model Release 10 min
OpenAI Cuts ChatGPT Limits for 1 Billion Free Users
OpenAI makes GPT-5.6 Luna free for ChatGPT's 1 billion weekly users, removing daily caps and adding a Think button for extended reasoning.
Jordan HaleAug 8, 2026
Model Release 12 min
ByteDance Builds 10T Model to Surpass Anthropic Mythos
ByteDance trains a 10T-parameter model targeting Anthropic Mythos, backed by $14B in Nvidia chips and a founder directive to stop distilling rivals.
Jordan HaleAug 8, 2026
Model Release 12 min
OpenAI Astra Beats Human Mathematicians on Ten Problems
OpenAI Astra solved ten long-standing math problems, from non-sofic groups to sphere packing, for just $2,000, challenging what research universities do.
Jordan HaleAug 5, 2026
Model Release 13 min
Alibaba Launches 2.4T-Parameter Model That Codes 16 Days
Alibaba's Qwen3.8-Max has 2.4 trillion parameters and a 1M-token context, completing a 16-day coding project solo. Open weights next week.
Jordan HaleAug 5, 2026
Model Release 12 min
Moonshot Kimi K3 Beats Fable 5 With Open-Weight Sparse MoE
Moonshot's 2.8T sparse MoE model ships with 1M token context, debuts #1 in coding benchmarks, and free weights arriving July 27 threaten frontier labs.
Jordan HaleJul 19, 2026
Model Release 13 min
OpenAI Sol Wins Commerce Clearance, Beats Anthropic
OpenAI GPT-5.6 Sol passes Commerce Department review at $30/M output tokens, 40% cheaper than Claude Opus 4.8, setting government clearance as a competitive moat.
Jordan HaleJul 14, 2026
Model Release 13 min
OpenAI GPT-5.6 Cuts Frontier Model Costs 67 Percent
OpenAI released three GPT-5.6 variants on July 9 with aggressive pricing ($0.60 Sol input tokens), forcing competitors to choose between margin defense and market share capture.
Jordan HaleJul 12, 2026
Model Release 13 min
Mistral Leanstral Cuts Formal Verification Costs 95 Percent
Mistral Leanstral 1.5 solves 587 of 672 PutnamBench proofs at $4 per problem, 75x cheaper than proprietary tools, finding five previously unknown bugs in open-source code.
Jordan HaleJul 12, 2026
Model Release 11 min
Anthropic Cuts Sonnet 5 to Opus Performance at Opus Price
Anthropic released Claude Sonnet 5 with Opus-level benchmarks at 20 percent lower cost, starting July 1, forcing competitors to respond.
Jordan HaleJul 2, 2026
Model Release 11 min
Google Gemini 3.5 Pro Doubles Context, Shifts AI Rivalry
Google's Gemini 3.5 Pro launches with 2M token context window, Deep Think reasoning mode, and $15/$60 pricing—the largest frontier model memory to date.
Jordan HaleJun 24, 2026
Model Release 11 min
Z.ai GLM-5.2 Builds Open Weights Governments Cannot Kill
Z.ai's GLM-5.2 ships a 1M-token open-weight model with an Anthropic-compatible API one day after the U.S. government banned Fable 5.
Jordan HaleJun 15, 2026
Model Release 13 min
Moonshot Kimi K2.7 Cuts Reasoning Tokens 30 Percent
Moonshot AI's Kimi K2.7 Code delivers 30% fewer reasoning tokens and 21.8% higher coding benchmark scores, open-sourced under MIT license on June 12.
Jordan HaleJun 13, 2026
Model Release 12 min
Anthropic Fable 5 Beats All AI Rivals on Code 2026
Anthropic's Fable 5 tops SWE-Bench Pro at 80.3%, beating GPT-5.5 by 22 points, opening the Mythos-class tier at $10 per million tokens.
Jordan HaleJun 10, 2026
Model Release 12 min
Apple Builds AFM Models With Google Gemini and Nvidia GPUs
Apple unveiled 5 AFM foundation models at WWDC 2026, distilling Google Gemini and running the cloud tier on Nvidia GPUs in Google Cloud infrastructure.
Jordan HaleJun 9, 2026
Model Release 14 min
OpenAI GPT-5.6 Leak Signals 1.5M Context June Launch 2026
OpenAI's GPT-5.6 appears in Codex routing logs under codename iris-alpha, with 1.5M tokens of context and 80-89% Polymarket odds of a June launch.
Jordan HaleJun 8, 2026
Model Release 12 min
OpenAI Kills GPT-4.5 to Signal the GPT-5 Era in 2026
OpenAI will retire GPT-4.5 from ChatGPT on June 27 and o3 on August 26, completing the first full clearout of the pre-GPT-5 model generation.
Jordan HaleJun 8, 2026
Model Release 12 min
Alibaba Qwen 3.7 Plus Launches Vision Agents at 6x Cut
Alibaba's Qwen 3.7 Plus adds vision and video to its Qwen backbone at $0.40 per million tokens, 6x cheaper than Qwen 3.7 Max, shipping June 1.
Jordan HaleJun 7, 2026
Model Release 11 min
Microsoft MAI-Thinking-1 Beats Claude Sonnet in 2026
Microsoft's MAI-Thinking-1, its first in-house reasoning model trained without OpenAI data, was preferred over Claude Sonnet 4.6 in blind human tests.
Jordan HaleJun 6, 2026
Model Release 12 min
Anthropic Sonnet 4.8 Leak Signals It Skips 4.7 in 2026
Anthropic's leaked 512,000-line Claude Code source map exposes Sonnet 4.8 and a skipped 4.7, plus hidden API tiers and routing thresholds rivals can study.
Jordan HaleJun 6, 2026
Model Release 12 min
Gemini 3.5 Pro Doubles Context to 2M Tokens in 2026
Gemini 3.5 Pro ships a 2 million token context window, double Flash, as Google launches its cheaper model first to pressure rivals on price.
Jordan HaleJun 6, 2026
Model Release 12 min
Microsoft MAI-Code-1-Flash Cuts Coding Tokens 60% 2026
Microsoft MAI-Code-1-Flash, a 5B in-house coding model, beats Claude Haiku 4.5 on SWE-Bench Pro and uses up to 60% fewer tokens in GitHub Copilot.
Jordan HaleJun 6, 2026
Model Release 12 min
Google Gemma 4 Launches a 12B Encoder Free Local Model
Google Gemma 4 12B is an encoder-free multimodal model that runs text, image, audio, and video on a 16GB laptop under Apache 2.0 with a 256K context.
Jordan HaleJun 6, 2026
Model Release 12 min
Microsoft MAI Launches 7 Models to Cut OpenAI Reliance
Microsoft launched 7 MAI models spanning code, reasoning, image, voice, and transcription at Build 2026, a full-stack bid to cut its OpenAI reliance.
Jordan HaleJun 6, 2026
Model Release 12 min
Claude Opus 4.8 Surpasses GPT-5.5 as the Top AI Model
Claude Opus 4.8 scores 61.4 on the Artificial Analysis Intelligence Index and 69.2% on SWE-Bench Pro, beating GPT-5.5 just 41 days after Opus 4.7.
Jordan HaleJun 5, 2026
Model Release 12 min
OpenAI Realtime 2 Cuts Voice AI Latency Below 1 Second
OpenAI's GPT-Realtime-2 cuts voice agent latency below one second with GPT-5-class reasoning and a 128K context, resetting voice AI economics.
Jordan HaleJun 5, 2026
Model Release 12 min
DeepSeek V4 Pro Undercuts US AI Coding Cost 95% 2026
DeepSeek V4-Pro made its 75% price cut permanent at $0.87 per million tokens, a frontier-grade coding model that undercuts US labs with open weights.
Jordan HaleJun 5, 2026
Model Release 12 min
MiniMax M3 Beats GPT-5.5 Coding at 1/12 Cost in 2026
MiniMax M3, an open-weight model from Shanghai, beats GPT-5.5 on SWE-Bench Pro at 1/12 the cost, putting the AI coding frontier up for free download.
Jordan HaleJun 5, 2026
Model Release 12 min
Ideogram 4 Launches 9.3B Open-Weight Image Model 2026
Ideogram 4 is a 9.3B open-weight text-to-image model with native 2K resolution that tops every open rival on text rendering and design quality.
Jordan HaleJun 5, 2026
Model Release 12 min
xAI Composer 2.5 Beats Grok 4.20 on Agentic Code 2026
xAI Composer 2.5 launched June 1, 2026, scoring 69.3 on agentic coding versus Grok 4.20 at 47.1, built on Moonshot Kimi K2.5 open weights.
Jordan HaleJun 5, 2026
Model Release 12 min
Qwen3.7 Max Surpasses Gemini Flash on AI Index 2026
Qwen3.7 Max scored 56.6 on the Artificial Analysis Index on May 20, 2026, the top Chinese model, beating DeepSeek V4 Pro on agentic coding.
Jordan HaleJun 5, 2026
Model Release 12 min
OpenAI Rosalind Cuts Genomics AI Compute 31% in 2026
OpenAI GPT-Rosalind beats GPT-5.5 across drug discovery and genomics while using 31% fewer tokens, and opens its science model preview worldwide.
Jordan HaleJun 5, 2026