Skip to content
AI Atlas
Diff the AI world

What changed between 5 Sept 2026 and 12 Sept 2026

Two dates, one scope: new and retired models, price and context changes, new benchmark leaders, new papers, provider and hardware changes. Built from the change log keyed on when things occurred — nothing is inferred, and the URL is the report.

Scope: the whole atlas

Reset

5,909 events in the window · 2,341 claims superseded · price rows opened 907 / closed 20 · entities at 5 Sept 2026: 0 → at 12 Sept 2026: 5,776. The entity lists are capped at 200 by the API — per-type counts below are “of the first 200”. Events are keyed on occurred_at (effective date when known) and exclude back-filled history unless include_backfill=1; new_entities uses first_seen_at. new_benchmark_leaders compares the primary-group leader computed from results observed by each date.

New models

New models 72

New models
ModelOrganizationParamsContextOpennessReleasedFirst seen
Phi-4-mini-instruct (3.8B), SigLIP2-so400M12 Sept 2026
Qwen2.5-OmniQwen Team7BOpen weights27 Mar 202512 Sept 2026
Qwen2.5-Omni-7BQwen Team12 Sept 2026
V4-FlashDeepSeek12 Sept 2026
qwen-mt-turboQwen TeamProprietary24 Jul 202512 Sept 2026
Document AI12 Sept 2026
Mistral OCR 312 Sept 2026
Mistral OCR 4Mistral AIProprietary23 Jun 202612 Sept 2026
deepseek-chat, deepseek-reasoner12 Sept 2026
DeepSeek-V4DeepSeek1.6T1MOpen weights12 Sept 2026
Qwen-VL12 Sept 2026
Qwen2-VL-7BQwen Team12 Sept 2026
Qwen2-VL-2BQwen Team12 Sept 2026
Qwen-Coder12 Sept 2026
Qwen2.5-Coder-Base12 Sept 2026
Qwen2.5-Coder-14BQwen Team12 Sept 2026
Qwen2.5-Coder-3BQwen Team12 Sept 2026
Qwen2.5-Coder-0.5BQwen Team12 Sept 2026
Vibe Enterprise12 Sept 2026
Vibe Team12 Sept 2026
Vibe Pro12 Sept 2026
Vibe12 Sept 2026
Qwen-TTSQwen TeamProprietary27 Jun 202512 Sept 2026
qwen-tts-2025-05-22Qwen Team12 Sept 2026
qwen-tts-latestQwen Team12 Sept 2026
Qwen3 Embedding SeriesQwen Team32.8KOpen weights5 Jun 202512 Sept 2026
Qwen3-Reranker-8BQwen Team12 Sept 2026
Qwen3-Reranker-4BQwen Team12 Sept 2026
Qwen3-Reranker-0.6BQwen Team12 Sept 2026
Qwen3-Embedding-8BQwen Team12 Sept 2026
Qwen3-Embedding-4BQwen Team12 Sept 2026
DeepSeek-R1-0528DeepSeekOpen weights12 Sept 2026
Qwen2.5-VL series12 Sept 2026
Qwen2.5-VL-32B-InstructQwen Team32BOpen weights24 Mar 202512 Sept 2026
Vision-language model specialized for grounding tasks such as pointing, counting, and object localization12 Sept 2026
Robostral NavigateMistral AI8BProprietary12 Sept 2026
Qwen2-VLQwen TeamOpen weights29 Aug 202412 Sept 2026
Qwen2.5-VL-3BQwen Team12 Sept 2026
Opus 4.812 Sept 2026
Qwen2.5 VLQwen TeamOpen weights26 Jan 202512 Sept 2026
Qwen VLoQwen Teamunknown26 Jun 202512 Sept 2026
Qwen2-Math12 Sept 2026
Qwen2.5-Math-RM-72BQwen Team12 Sept 2026
Qwen2.5-Math-7B-InstructQwen Team12 Sept 2026
Qwen-Audio12 Sept 2026
Qwen12 Sept 2026
Qwen2-AudioQwen Team7BOpen weights9 Aug 202412 Sept 2026
Qwen2-Audio-7B-InstructQwen Team12 Sept 2026
Qwen2-Audio-7BQwen Team12 Sept 2026
QwQ-Max-PreviewQwen TeamOpen weights25 Feb 202512 Sept 2026
Qwen-Image-EditQwen Team20Bunknown19 Aug 202512 Sept 2026
CodeQwen1.512 Sept 2026
Qwen2.5-Coder-InstructQwen Team12 Sept 2026
V3.1-Terminus12 Sept 2026
V3.2-Exp12 Sept 2026
QVQ-72B-PreviewQwen Team72BOpen weights25 Dec 202412 Sept 2026
QVQ-MaxQwen TeamProprietary28 Mar 202512 Sept 2026
Qwen2.5-1MQwen Team1MOpen weights27 Jan 202512 Sept 2026
Qwen2.5-14B-Instruct-1MQwen Team12 Sept 2026
Qwen2.5-7B-Instruct-1MQwen Team12 Sept 2026
Qwen212 Sept 2026
Qwen2-VL-72BQwen Team12 Sept 2026
Qwen2.5-Math-1.5B-InstructQwen Team12 Sept 2026
Qwen2.5-Math-1.5BQwen Team12 Sept 2026
Qwen2.5-Math-72B-InstructQwen Team12 Sept 2026
Qwen2.5-Math-72BQwen Team12 Sept 2026
Qwen2.5-Coder-1.5BQwen Team12 Sept 2026
Qwen2.5-14B-InstructQwen Team12 Sept 2026
Qwen2.5-14BQwen Team12 Sept 2026
Qwen2.5-MathQwen Team4.1KOpen weights19 Sept 202412 Sept 2026
Qwen2.5-CoderQwen Team32.5B128Krestricted-weights12 Nov 202412 Sept 2026
Qwen2.5Qwen Team72.7B131.1Krestricted-weights19 Sept 202412 Sept 2026

Retired models

Retired models 1

Price changes

Price changes 20

Price changes
DateModelProviderInput / 1MOutput / 1MSource
DeepSeek V4 Flash (0731)DeepSeek API$0.065$0.04−38.5%$0.18$0.08−55.6%openrouter.ai
DeepSeek V4 Pro 0423DeepSeek API$0.829$0.819−1.2%$1.66$1.64−1.2%openrouter.ai
DeepSeek V4 Flash 0423DeepSeek API$0.067$0.067−0.4%$0.134$0.134−0.4%openrouter.ai
DeepSeek V4 Flash 0423DeepSeek API$0.067$0.067+0.4%$0.134$0.134+0.4%openrouter.ai
DeepSeek V4 Pro 0423DeepSeek API$0.85$0.829−2.4%$1.7$1.66−2.4%openrouter.ai
DeepSeek V4 Flash 0423DeepSeek API$0.067$0.067−0.8%$0.135$0.134−0.8%openrouter.ai
Qwen3.8 27BOpenRouter$0.42$0.214−49.0%$3$2.55−15.0%openrouter.ai
DeepSeek-V4-Pro-0813DeepSeek API$0.579$0.578−0.2%$1.74$1.73−0.2%openrouter.ai
Kimi K3Moonshot AI Platform$1.54$2.3+50.0%$7.7$11.55+50.0%openrouter.ai
DeepSeek V4 Pro 0423DeepSeek API$0.84$0.85+1.2%$1.68$1.7+1.2%openrouter.ai
DeepSeek V4 Flash 0423DeepSeek API$0.068$0.067−0.4%$0.135$0.135−0.4%openrouter.ai
MiniMax M2.5MiniMax API$0.27$0.30+11.1%$1.08$1.2+11.1%openrouter.ai
Qwen3 235B A22B Instruct 2507OpenRouter$0.22$0.087−60.2%$0.88$0.35−60.2%openrouter.ai
Llama 3.1 70B InstructOpenRouter$0.40$0.72+80.0%$0.40$0.72+80.0%openrouter.ai
Kimi K3Moonshot AI Platform$1.8$1.54−14.5%$9.01$7.7−14.5%openrouter.ai
Hy3OpenRouter$0.083$0.132+60.0%$0.33$0.528+60.0%openrouter.ai
DeepSeek V4 Pro 0423DeepSeek API$0.948$0.84−11.4%$1.9$1.68−11.4%openrouter.ai
DeepSeek V4 Flash 0423DeepSeek API$0.085$0.068−20.0%$0.169$0.135−20.0%openrouter.ai
DeepSeek V4 Pro 0423DeepSeek API$0.86$0.948+10.2%$1.72$1.9+10.2%openrouter.ai
DeepSeek V4 Flash 0423DeepSeek API$0.085$0.085−0.3%$0.17$0.169−0.3%openrouter.ai

Context changes

Context changes 103

Context changes
DateModelOrganizationContext windowSourceHistory
Mistral Small 4Mistral AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
Z.ai GLM 5.2Z.ai (Zhipu AI)1M tokens1.05M tokens+4.9%openrouter.ai claims →
gpt-5.6-solOpenAI1M tokens1.05M tokens+5.0%openrouter.ai claims →
gpt-5.6-terraOpenAI1M tokens1.05M tokens+5.0%openrouter.ai claims →
gpt-5.6-lunaOpenAI1M tokens1.05M tokens+5.0%openrouter.ai claims →
gpt-6-astraOpenAI1M tokens1.05M tokens+5.0%openrouter.ai claims →
Gemini 3.1 Flash-Lite PreviewGoogle1M tokens1.05M tokens+4.9%ai.google.dev claims →
Gemini 2.5 Flash-LiteGoogle1M tokens1.05M tokens+4.9%ai.google.dev claims →
Gemini 2.5 FlashGoogle1M tokens1.05M tokens+4.9%ai.google.dev claims →
Gemini 3.5 FlashGoogle1M tokens1.05M tokens+4.9%ai.google.dev claims →
Gemini 3.7 FlashGoogle1M tokens1.05M tokens+4.9%ai.google.dev claims →
Gemini 3.8 FlashGoogle1M tokens1.05M tokens+4.9%ai.google.dev claims →
Qwen3 235B A22BQwen128K tokens131.1K tokens+2.4%openrouter.ai claims →
Kimi K2.7 CodeMoonshot AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
Qwen3.6 Max PreviewQwen256K tokens262.1K tokens+2.4%openrouter.ai claims →
gpt-5.5-proOpenAI922K tokens1.05M tokens+13.9%openrouter.ai claims →
gpt-5.5OpenAI922K tokens1.05M tokens+13.9%openrouter.ai claims →
Hy3 previewTencent256K tokens262.1K tokens+2.4%openrouter.ai claims →
MiMo-V2.5-ProXiaomi1M tokens1.05M tokens+5.0%openrouter.ai claims →
MiMo-V2.5Xiaomi1M tokens1.05M tokens+5.0%openrouter.ai claims →
Kimi K2.6Moonshot AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
GLM 5.1Z.ai (Zhipu AI)200K tokens204.8K tokens+2.4%openrouter.ai claims →
Gemma 4 26B A4BGoogle256K tokens262.1K tokens+2.4%openrouter.ai claims →
Gemma 4 31BGoogle256K tokens262.1K tokens+2.4%openrouter.ai claims →
GLM 5V TurboZ.ai (Zhipu AI)200K tokens202.8K tokens+1.4%openrouter.ai claims →
Trinity Large ThinkingArcee AI512K tokens262.1K tokens−48.8%openrouter.ai claims →
KAT-Coder-Pro V2Kwaipilot256K tokens262.1K tokens+2.4%openrouter.ai claims →
GLM 5 TurboZ.ai (Zhipu AI)200K tokens202.8K tokens+1.4%openrouter.ai claims →
Nemotron 3 SuperNVIDIA1M tokens262.1K tokens−73.8%openrouter.ai claims →
GLM 5Z.ai (Zhipu AI)200K tokens204.8K tokens+2.4%openrouter.ai claims →
Qwen3 Coder NextQwen256K tokens262.1K tokens+2.4%openrouter.ai claims →
Step 3.5 FlashStepFun256K tokens262.1K tokens+2.4%openrouter.ai claims →
Kimi K2.5Moonshot AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
Solar Pro 3Upstage128K tokens131.1K tokens+2.4%openrouter.ai claims →
GLM 4.7Z.ai (Zhipu AI)200K tokens204.8K tokens+2.4%openrouter.ai claims →
Nemotron 3 Nano 30B A3BNVIDIA1M tokens262.1K tokens−73.8%openrouter.ai claims →
Devstral 2Mistral AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
GLM 4.6VZ.ai (Zhipu AI)128K tokens131.1K tokens+2.4%openrouter.ai claims →
DeepSeek V3.2DeepSeek128K tokens163.8K tokens+28.0%openrouter.ai claims →
gpt-5.1OpenAI272K tokens400K tokens+47.1%openrouter.ai claims →
Kimi K2 ThinkingMoonshot AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
Qwen3 VL 32B InstructQwen256K tokens131.1K tokens−48.8%openrouter.ai claims →
Granite 4.0 MicroIBM128K tokens131K tokens+2.3%openrouter.ai claims →
Qwen3 VL 8B InstructQwen256K tokens262.1K tokens+2.4%openrouter.ai claims →
Qwen3 VL 30B A3B InstructQwen256K tokens262.1K tokens+2.4%openrouter.ai claims →
GLM 4.6Z.ai (Zhipu AI)200K tokens204.8K tokens+2.4%openrouter.ai claims →
DeepSeek V3.2 ExpDeepSeek128K tokens163.8K tokens+28.0%openrouter.ai claims →
DeepSeek V3.1 TerminusDeepSeek128K tokens163.8K tokens+28.0%openrouter.ai claims →
Kimi K2 0905Moonshot AI256K tokens262.1K tokens+2.4%openrouter.ai claims →
Hermes 4 405BNous Research128K tokens131.1K tokens+2.4%openrouter.ai claims →
DeepSeek V3.1DeepSeek128K tokens163.8K tokens+28.0%openrouter.ai claims →
Mistral Medium 3.1Mistral AI128K tokens131.1K tokens+2.4%openrouter.ai claims →
GLM 4.5VZ.ai (Zhipu AI)64K tokens65.5K tokens+2.4%openrouter.ai claims →
GLM 4.5Z.ai (Zhipu AI)128K tokens131.1K tokens+2.4%openrouter.ai claims →
GLM 4.5 AirZ.ai (Zhipu AI)128K tokens131.1K tokens+2.4%openrouter.ai claims →
Qwen3 235B A22B Instruct 2507Qwen256K tokens262.1K tokens+2.4%openrouter.ai claims →
Kimi K2 0711Moonshot AI128K tokens131.1K tokens+2.4%openrouter.ai claims →
Mistral Medium 3Mistral AI128K tokens131.1K tokens+2.4%openrouter.ai claims →
Qwen3 14BQwen32.8K tokens131.1K tokens+300.0%openrouter.ai claims →
Qwen3 32BQwen32.8K tokens131.1K tokens+300.0%openrouter.ai claims →
gpt-4.1OpenAI1M tokens1.05M tokens+4.8%openrouter.ai claims →
gpt-4.1-miniOpenAI1M tokens1.05M tokens+4.8%openrouter.ai claims →
gpt-4.1-nanoOpenAI1M tokens1.05M tokens+4.8%openrouter.ai claims →
Llama 4 MaverickMeta AI1M tokens1.05M tokens+4.9%openrouter.ai claims →
Llama 4 ScoutMeta AI10M tokens1.31M tokens−86.9%openrouter.ai claims →
DeepSeek V3 0324DeepSeek128K tokens163.8K tokens+28.0%openrouter.ai claims →
Gemma 3 4BGoogle128K tokens131.1K tokens+2.4%openrouter.ai claims →
Gemma 3 12BGoogle128K tokens131.1K tokens+2.4%openrouter.ai claims →
Reka Flash 3rekaai128K tokens65.5K tokens−48.8%openrouter.ai claims →
Gemma 3 27BGoogle128K tokens131.1K tokens+2.4%openrouter.ai claims →
Sonar Reasoning ProPerplexity AI127K tokens128K tokens+0.8%openrouter.ai claims →
Mistral SabaMistral AI32K tokens32.8K tokens+2.4%openrouter.ai claims →
Mistral Small 3Mistral AI32K tokens32.8K tokens+2.4%openrouter.ai claims →
SonarPerplexity AI127K tokens127.1K tokens+0.1%openrouter.ai claims →
R1 Distill Llama 70BDeepSeek128K tokens8.19K tokens−93.6%openrouter.ai claims →
Phi 4Microsoft16K tokens16.4K tokens+2.4%openrouter.ai claims →
DeepSeek V3DeepSeek128K tokens163.8K tokens+28.0%openrouter.ai claims →
Mistral Large 2.0Mistral AI128K tokens131.1K tokens+2.4%openrouter.ai claims →
Qwen2.5 Coder 32B InstructQwen131.1K tokens32.8K tokens−75.0%openrouter.ai claims →
Hermes 3 70B InstructNous Research128K tokens131.1K tokens+2.4%openrouter.ai claims →
Mistral LargeMistral AI32.8K tokens128K tokens+290.6%openrouter.ai claims →
gpt-3.5-turboOpenAI4.1K tokens4.1K tokens−0.0%openrouter.ai claims →
gpt-4OpenAI8.19K tokens8.19K tokens−0.0%openrouter.ai claims →
NVIDIA Nemotron 3.5 Lightning 30B A3BNVIDIA1M tokens262.1K tokens−73.8%openrouter.ai claims →
Ling 3.0 Flash VLinclusionAI262.1K tokens131.1K tokens−50.0%openrouter.ai claims →
Muse Spark 1.3Meta AI1M tokens1.05M tokens+4.9%openrouter.ai claims →
Qwen3.8 FlashQwen256K tokens1M tokens+290.6%openrouter.ai claims →
GLM 5.3 FlashZ.ai (Zhipu AI)1M tokens1.31M tokens+31.1%openrouter.ai claims →
GLM 5.3Z.ai (Zhipu AI)1M tokens1.31M tokens+31.1%openrouter.ai claims →
Qwen3.8 27BQwen256K tokens1M tokens+290.6%openrouter.ai claims →
Qwen3.8 2.4T A95BQwen983.6K tokens1.05M tokens+6.6%openrouter.ai claims →
Solar Pro 4Upstage512K tokens524.3K tokens+2.4%openrouter.ai claims →
Inkling SmallThinking Machines1M tokens1.05M tokens+4.9%openrouter.ai claims →
LongCat 2.0Meituan1M tokens1.05M tokens+4.9%openrouter.ai claims →
InklingThinking Machines1M tokens1.05M tokens+4.9%openrouter.ai claims →
Hy3Tencent256K tokens262.1K tokens+2.4%openrouter.ai claims →
gpt-3.5-turboOpenAI4.1K tokens16.4K tokens+300.0%openrouter.ai claims →
MiniMax-M3MiniMax1M tokens1.05M tokens+4.9%openrouter.ai claims →
Qwen3 MaxQwen256K tokens262.1K tokens+2.4%openrouter.ai claims →
Qwen2.5 72B InstructQwen131.1K tokens32.8K tokens−75.0%openrouter.ai claims →
Qwen3 235B A22B Instruct 2507Qwen16.4K tokens235.9K tokens+1340.0%openrouter.ai claims →
Llama 3.1 70B InstructMeta AI16.4K tokens8.19K tokens−50.0%openrouter.ai claims →
MiniMax M2.5MiniMax128K tokens131.1K tokens+2.4%openrouter.ai claims →

New benchmark leaders

New benchmark leaders 23

Benchmarks whose primary-group leader (computed from results observed by each date) differs between 5 Sept 2026 and 12 Sept 2026.

New benchmark leaders
BenchmarkLeader at 5 Sept 2026ScoreLeader at 12 Sept 2026ScoreMetric
Aider polyglotcodingno result observed yetgpt-586.7pass_rate_2official-benchmark
Aider polyglot — well-formed responsescodingno result observed yetCodestral 25.01100percent_cases_well_formedofficial-benchmark
Artificial Analysis Intelligence Indexcompositeno result observed yetClaude Fable 5.153.37indexindependent-evaluator
GPQA Diamondreasoningno result observed yetgpt-6-astra96.26accuracyindependent-evaluator
Humanity's Last Examknowledgeno result observed yetClaude Fable 5.159.13accuracyindependent-evaluator
IFBenchinstruction-followingno result observed yetGrok 4.383.33accuracyindependent-evaluator
LiveBenchgeneralno result observed yetClaude Fable 5.183.41global_averageofficial-benchmark
LiveBench Agentic Codingcodingno result observed yetDeepSeek-V4.1-Flash77.27average scoreofficial-benchmark
LiveBench Codingcodingno result observed yetClaude Fable 5.186.38average scoreofficial-benchmark
LiveBench Data Analysisgeneralno result observed yetgpt-6-astra82.97average scoreofficial-benchmark
LiveBench Instruction Followinginstruction-followingno result observed yetGemini 3.8 Flash81.41average scoreofficial-benchmark
LiveBench Languagegeneralno result observed yetClaude Fable 590.68average scoreofficial-benchmark
LiveBench Mathematicsmathno result observed yetClaude Fable 5.197.01average scoreofficial-benchmark
LiveBench Reasoningreasoningno result observed yetgpt-6-astra92.65average scoreofficial-benchmark
MMMU-Promultimodalno result observed yetgpt-6-astra86.88accuracyindependent-evaluator
SWE-bench (full test split)codingno result observed yetclaude-3-opus3.79resolvedofficial-benchmark
SWE-bench Litecodingno result observed yetclaude-3-opus4.33resolvedofficial-benchmark
SWE-bench Multilingualcodingno result observed yetgemini-3-flash72.7resolvedofficial-benchmark
SWE-bench Multimodalcodingno result observed yeto335.98resolvedofficial-benchmark
SWE-bench Verifiedcodingno result observed yetClaude Opus 4.576.8resolvedofficial-benchmark
SciCodecodingno result observed yetClaude Fable 5.163.08accuracyindependent-evaluator
Terminal-Benchagenticno result observed yetgpt-5.6-sol65.91accuracyindependent-evaluator
τ²-benchagenticno result observed yetZ.ai GLM 5.299.12pass^1independent-evaluator

New papers

New papers 21

New papers
PaperOrganizationPublishedFirst seen
A Multi-Modal Perception Pipeline for Object Detection and Tracking in Autonomous Racing12 Sept 202612 Sept 2026
Your Agent Says Yes: Interpreting Adversarial Market Behavior Beyond Individual Transactions12 Sept 202612 Sept 2026
Monadic Second-Order Logic in HOL: Deep and Shallow with Automated Faithfulness (Extended Preprint)12 Sept 202612 Sept 2026
Ordinary, Reasonable Chatbots: Do AI Models Track Human Legal Judgments?12 Sept 202612 Sept 2026
Toward Collective-Centric Evaluation of Preference Inference for Participatory Democracy12 Sept 202612 Sept 2026
OmegaUse-SOP: SOP Engineering for Professional Computer Use from Human Demonstrations12 Sept 202612 Sept 2026
FaultLens: Learning Compact Behavioral Test Suites for Generated Operational Programs12 Sept 202612 Sept 2026
GitSkills: A Dataset of Agent Skills on GitHub12 Sept 202612 Sept 2026
Natural Language Access to Domain-Specific Metadata: A Reusable Framework for LLM Query Generation12 Sept 202612 Sept 2026
Adaptive Perturbation Selection for Contrastive Audio Decoding12 Sept 202612 Sept 2026
From Agent Traces to Trust: A Survey of Evidence Tracing and Execution Provenance in LLM Agents12 Sept 202612 Sept 2026
Causal Past Logic for Runtime Verification of Distributed LLM Agent Workflows12 Sept 202612 Sept 2026
Four Generations of Quantum Biomedical Sensors12 Sept 202612 Sept 2026
DCO: Dynamic Cache Orchestration for LLM Accelerators through Predictive Management12 Sept 202612 Sept 2026
Generative AI performance in core undergraduate mathematics: a curriculum-level case study12 Sept 202612 Sept 2026
Amulet: a Python Library for Assessing Interactions Among ML Defenses and Risks12 Sept 202612 Sept 2026
Spectral Masking and Interpolation Attack (SMIA): A Black-box Adversarial Attack against Voice Authentication and Anti-Spoofing Systems12 Sept 202612 Sept 2026
A Survey of Threats Against Voice Authentication and Anti-Spoofing Systems12 Sept 202612 Sept 2026
Exploring Multimodal Prompt for Visualization Authoring with Large Language Models12 Sept 202612 Sept 2026
GoAnt: Quality-Diversity Multi-Agent Search for Alpha Factor Discovery in Market Microstructure Data12 Sept 202612 Sept 2026
FastE: Readout-Triggered Token Compression for LLM Embedding Inference12 Sept 202612 Sept 2026

Provider changes

Provider changes 283

Provider changes
DateModelProviderInput / 1MOutput / 1MSource
Mistral AIMistral AImistral.ai
SonarOpenRouter$1$1openrouter.ai
gpt-4oOpenRouter$2.5$10openrouter.ai
GPT-4o (2024-05-13)OpenRouter$5$15openrouter.ai
gpt-4oOpenRouter$1.25$5openrouter.ai
Mixtral 8x22B InstructOpenRouter$2$6openrouter.ai
gpt-4-turboOpenRouter$10$30openrouter.ai
gpt-4-turboOpenRouter$5$15openrouter.ai
Claude 3 HaikuOpenRouter$0.25$1.25openrouter.ai
Mistral LargeOpenRouter$2$6openrouter.ai
gpt-3.5-turboOpenRouter$1$2openrouter.ai
GPT-4 Turbo PreviewOpenRouter$10$30openrouter.ai
gpt-3.5-turbo-instructOpenRouter$1.5$2openrouter.ai
gpt-3.5-turbo-16kOpenRouter$3$4openrouter.ai
gpt-3.5-turboOpenRouter$0.50$1.5openrouter.ai
gpt-3.5-turboOpenRouter$0.25$0.75openrouter.ai
gpt-4OpenRouter$30$60openrouter.ai
DeepSeek-V4.1-FlashOpenRouter$0.15$0.60openrouter.ai
gpt-6-astraOpenRouter$10$50openrouter.ai
gpt-6-astraOpenRouter$5$25openrouter.ai
GPT-6 Astra ProOpenRouter$10$50openrouter.ai
GPT-6 Astra ProOpenRouter$5$25openrouter.ai
Gemini 3.8 FlashOpenRouter$0.75$3.75openrouter.ai
Gemini 3.8 FlashOpenRouter$0.375$1.88openrouter.ai
Claude Fable 5.1OpenRouter$10$50openrouter.ai
Claude Fable 5.1OpenRouter$5$25openrouter.ai
GLM 5.3 FlashOpenRouter$0.15$0.50openrouter.ai
GLM 5.3 FlashOpenRouter$0.075$0.25openrouter.ai
DeepSeek V4 Flash Vision ExpOpenRouter$0.22$0.66openrouter.ai
DeepSeek V4 Flash Vision ExpOpenRouter$0.11$0.33openrouter.ai
GLM 5.3OpenRouter$1.4$4.4openrouter.ai
GLM 5.3OpenRouter$0.70$2.2openrouter.ai
Gemini 3.7 FlashOpenRouter$0.75$3.75openrouter.ai
Gemini 3.7 FlashOpenRouter$0.375$1.88openrouter.ai
DeepSeek-V4-Pro-0813OpenRouter$0.578$1.73openrouter.ai
DeepSeek-V4-Pro-0813OpenRouter$0.66$1.98openrouter.ai
Grok 4.6OpenRouter$2$6openrouter.ai
NVIDIA Nemotron 3.5 Lightning 30B A3BOpenRouter$0.08$0.20openrouter.ai
NVIDIA Nemotron 3.5 Lightning 30B A3BOpenRouter$0$0openrouter.ai
DeepSeek V4 Flash (0731)OpenRouter$0.04$0.08openrouter.ai
DeepSeek V4 Flash (0731)OpenRouter$0.11$0.33openrouter.ai
Claude Opus 5OpenRouter$5$25openrouter.ai
Claude Opus 5OpenRouter$2.5$12.5openrouter.ai
Gemini 3.6 FlashOpenRouter$0.75$3.75openrouter.ai
Gemini 3.6 FlashOpenRouter$0.375$1.88openrouter.ai
Gemini 3.5 Flash-LiteOpenRouter$0.30$2.5openrouter.ai
Gemini 3.5 Flash-LiteOpenRouter$0.15$1.25openrouter.ai
Kimi K3OpenRouter$2.3$11.55openrouter.ai
Kimi K3OpenRouter$3$15openrouter.ai
GPT-5.6 Luna ProOpenRouter$0.20$1.2openrouter.ai
GPT-5.6 Luna ProOpenRouter$0.10$0.60openrouter.ai
gpt-5.6-lunaOpenRouter$0.20$1.2openrouter.ai
gpt-5.6-lunaOpenRouter$0.10$0.60openrouter.ai
GPT-5.6 Terra ProOpenRouter$2$12openrouter.ai
GPT-5.6 Terra ProOpenRouter$1$6openrouter.ai
gpt-5.6-terraOpenRouter$2$12openrouter.ai
gpt-5.6-terraOpenRouter$1$6openrouter.ai
GPT-5.6 Sol ProOpenRouter$2$10openrouter.ai
GPT-5.6 Sol ProOpenRouter$1$5openrouter.ai
gpt-5.6-solOpenRouter$2$10openrouter.ai
gpt-5.6-solOpenRouter$1$5openrouter.ai
Grok 4.5OpenRouter$2$6openrouter.ai
Claude Sonnet 5OpenRouter$2$10openrouter.ai
Claude Sonnet 5OpenRouter$1$5openrouter.ai
Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)OpenRouter$0.25$1.5openrouter.ai
Gemini 3.1 Flash Image (Nano Banana 2)OpenRouter$0.50$3openrouter.ai
Gemini 3 Pro Image (Nano Banana Pro)OpenRouter$2$12openrouter.ai
North Mini Code (free)OpenRouter$0$0openrouter.ai
Z.ai GLM 5.2OpenRouter$0.60$2openrouter.ai
Z.ai GLM 5.2OpenRouter$0.70$2.2openrouter.ai
Kimi K2.7 CodeOpenRouter$0.71$3.5openrouter.ai
Claude Fable 5OpenRouter$10$50openrouter.ai
Claude Fable 5OpenRouter$5$25openrouter.ai
Nemotron 3.5 Content SafetyOpenRouter$0.20$0.20openrouter.ai
Nemotron 3.5 Content SafetyOpenRouter$0$0openrouter.ai
Nemotron 3 UltraOpenRouter$0.625$3.13openrouter.ai
Nemotron 3 UltraOpenRouter$0$0openrouter.ai
MiniMax-M3OpenRouter$0.30$1.2openrouter.ai
MiniMax-M3OpenRouter$0.30$1.2openrouter.ai
Claude Opus 4.8OpenRouter$5$25openrouter.ai
Claude Opus 4.8OpenRouter$2.5$12.5openrouter.ai
Grok Build 0.1OpenRouter$1$2openrouter.ai
Gemini 3.5 FlashOpenRouter$1.5$9openrouter.ai
Gemini 3.5 FlashOpenRouter$0.75$4.5openrouter.ai
Gemini 3.1 Flash-LiteOpenRouter$0.25$1.5openrouter.ai
Gemini 3.1 Flash-LiteOpenRouter$0.125$0.75openrouter.ai
GPT Chat LatestOpenRouter$5$30openrouter.ai
Grok 4.3OpenRouter$1.25$2.5openrouter.ai
Grok 4.3OpenRouter$1$2openrouter.ai
Mistral Medium 3.5OpenRouter$1.5$7.5openrouter.ai
Mistral Medium 3.5OpenRouter$0.75$3.75openrouter.ai
Nemotron 3 Nano Omni (free)OpenRouter$0$0openrouter.ai
gpt-5.5-proOpenRouter$30$180openrouter.ai
gpt-5.5-proOpenRouter$15$90openrouter.ai
gpt-5.5OpenRouter$5$30openrouter.ai
gpt-5.5OpenRouter$2.5$15openrouter.ai
DeepSeek V4 Pro 0423OpenRouter$0.809$1.62openrouter.ai
DeepSeek V4 Flash 0423OpenRouter$0.067$0.134openrouter.ai
GPT-5.4 Image 2OpenRouter$8$15openrouter.ai
Kimi K2.6OpenRouter$0.95$4openrouter.ai
Claude Opus 4.7OpenRouter$5$25openrouter.ai
Claude Opus 4.7OpenRouter$2.5$12.5openrouter.ai
GLM 5.1OpenRouter$0.966$3.04openrouter.ai
Gemma 4 26B A4BOpenRouter$0.042$0.22openrouter.ai
Gemma 4 26B A4BOpenRouter$0$0openrouter.ai
Gemma 4 31BOpenRouter$0.09$0.34openrouter.ai
Gemma 4 31BOpenRouter$0.39$0.97openrouter.ai
Gemma 4 31BOpenRouter$0$0openrouter.ai
GLM 5V TurboOpenRouter$1.2$4openrouter.ai
Grok 4.20 Multi-AgentOpenRouter$1.25$2.5openrouter.ai
Grok 4.20OpenRouter$1.25$2.5openrouter.ai
Lyria 3 Pro PreviewOpenRouter$0$0openrouter.ai
Lyria 3OpenRouter$0$0openrouter.ai
MiniMax M2.7OpenRouter$0.30$1.2openrouter.ai
gpt-5.4-nanoOpenRouter$0.20$1.25openrouter.ai
gpt-5.4-nanoOpenRouter$0.10$0.625openrouter.ai
gpt-5.4-miniOpenRouter$0.75$4.5openrouter.ai
gpt-5.4-miniOpenRouter$0.375$2.25openrouter.ai
Mistral Small 4OpenRouter$0.15$0.60openrouter.ai
Mistral Small 4OpenRouter$0.075$0.30openrouter.ai
GLM 5 TurboOpenRouter$1.2$4openrouter.ai
Nemotron 3 SuperOpenRouter$0.085$0.40openrouter.ai
Nemotron 3 SuperOpenRouter$0$0openrouter.ai
gpt-5.4-proOpenRouter$30$180openrouter.ai
gpt-5.4-proOpenRouter$15$90openrouter.ai
gpt-5.4OpenRouter$2.5$15openrouter.ai
gpt-5.4OpenRouter$1.25$7.5openrouter.ai
Gemini 3.1 Flash-Lite PreviewOpenRouter$0.25$1.5openrouter.ai
Nano Banana 2 (Gemini 3.1 Flash Image Preview)OpenRouter$0.50$3openrouter.ai
Gemini 3.1 Pro Preview Custom ToolsOpenRouter$2$12openrouter.ai
gpt-5.3-codexOpenRouter$1.75$14openrouter.ai
Gemini 3.1 Pro PreviewOpenRouter$2$12openrouter.ai
Gemini 3.1 Pro PreviewOpenRouter$1$6openrouter.ai
Claude Sonnet 4.6OpenRouter$3$15openrouter.ai
Claude Sonnet 4.6OpenRouter$1.5$7.5openrouter.ai
MiniMax M2.5OpenRouter$0.30$1.2openrouter.ai
GLM 5OpenRouter$0.60$1.92openrouter.ai
Claude Opus 4.6OpenRouter$5$25openrouter.ai
Claude Opus 4.6OpenRouter$2.5$12.5openrouter.ai
Kimi K2.5OpenRouter$0.45$2.25openrouter.ai
MiniMax M2-herOpenRouter$0.30$1.2openrouter.ai
gpt-audioOpenRouter$2.5$10openrouter.ai
gpt-audio-miniOpenRouter$0.60$2.4openrouter.ai
GLM 4.7 FlashOpenRouter$0.06$0.40openrouter.ai
GPT-5.2-CodexOpenRouter$1.75$14openrouter.ai
MiniMax M2.1OpenRouter$0.30$1.2openrouter.ai
GLM 4.7OpenRouter$0.40$1.75openrouter.ai
Gemini 3 Flash PreviewOpenRouter$0.50$3openrouter.ai
Gemini 3 Flash PreviewOpenRouter$0.25$1.5openrouter.ai
Nemotron 3 Nano 30B A3BOpenRouter$0.05$0.20openrouter.ai
GPT-5.2 ChatOpenRouter$1.75$14openrouter.ai
gpt-5.2-proOpenRouter$21$168openrouter.ai
gpt-5.2-proOpenRouter$10.5$84openrouter.ai
gpt-5.2OpenRouter$1.75$14openrouter.ai
gpt-5.2OpenRouter$0.875$7openrouter.ai
Devstral 2OpenRouter$0.40$2openrouter.ai
GLM 4.6VOpenRouter$0.30$0.90openrouter.ai
GPT-5.1-Codex-MaxOpenRouter$1.25$10openrouter.ai
Ministral 3 14B 2512OpenRouter$0.20$0.20openrouter.ai
Ministral 3 8B 2512OpenRouter$0.15$0.15openrouter.ai
Ministral 3 8B 2512OpenRouter$0.075$0.075openrouter.ai
Ministral 3 3B 2512OpenRouter$0.10$0.10openrouter.ai
Mistral Large 3 2512OpenRouter$0.50$1.5openrouter.ai
Mistral Large 3 2512OpenRouter$0.25$0.75openrouter.ai
DeepSeek V3.2OpenRouter$0.269$0.40openrouter.ai
Claude Opus 4.5OpenRouter$5$25openrouter.ai
Claude Opus 4.5OpenRouter$2.5$12.5openrouter.ai
Nano Banana Pro (Gemini 3 Pro Image Preview)OpenRouter$2$12openrouter.ai
gpt-5.1OpenRouter$1.25$10openrouter.ai
gpt-5.1OpenRouter$0.625$5openrouter.ai
GPT-5.1-CodexOpenRouter$1.25$10openrouter.ai
GPT-5.1-Codex MiniOpenRouter$0.25$2openrouter.ai
Kimi K2 ThinkingOpenRouter$0.60$2.5openrouter.ai
Sonar Pro SearchOpenRouter$3$15openrouter.ai
Voxtral Small 24B 2507OpenRouter$0.10$0.30openrouter.ai
gpt-oss-safeguard-20bOpenRouter$0.075$0.30openrouter.ai
MiniMax M2OpenRouter$0.255$1.02openrouter.ai
GPT-5 Image MiniOpenRouter$2.5$2openrouter.ai
Claude Haiku 4.5OpenRouter$1$5openrouter.ai
Claude Haiku 4.5OpenRouter$0.50$2.5openrouter.ai
GPT-5 ImageOpenRouter$10$10openrouter.ai
Gemini 2.5 Flash Image (Nano Banana)OpenRouter$0.30$2.5openrouter.ai
gpt-5-proOpenRouter$15$120openrouter.ai
gpt-5-proOpenRouter$7.5$60openrouter.ai
GLM 4.6OpenRouter$0.43$1.75openrouter.ai
Claude Sonnet 4.5OpenRouter$3$15openrouter.ai
Claude Sonnet 4.5OpenRouter$1.5$7.5openrouter.ai
DeepSeek V3.2 ExpOpenRouter$0.27$0.41openrouter.ai
DeepSeek V3.1 TerminusOpenRouter$0.27$1openrouter.ai
Kimi K2 0905OpenRouter$0.60$2.5openrouter.ai
DeepSeek V3.1OpenRouter$0.25$0.95openrouter.ai
Mistral Medium 3.1OpenRouter$0.40$2openrouter.ai
Mistral Medium 3.1OpenRouter$0.20$1openrouter.ai
GLM 4.5VOpenRouter$0.60$1.8openrouter.ai
gpt-5OpenRouter$1.25$10openrouter.ai
gpt-5OpenRouter$0.625$5openrouter.ai
gpt-5-miniOpenRouter$0.25$2openrouter.ai
gpt-5-miniOpenRouter$0.125$1openrouter.ai
gpt-5-nanoOpenRouter$0.05$0.40openrouter.ai
gpt-5-nanoOpenRouter$0.025$0.20openrouter.ai

Showing the first 200 of 283 price events. Narrow the scope or the dates to see everything.

Hardware changes

Hardware changes 0

No hardware change between these dates

Other property changes

Other property changes 200

Openness, licence, status, parameters, release-date and other material property changes (context changes are listed above).

  • Openness changedModelNorth Small TranslateCohere

    North Small Translate: openness changed from restricted-weights to open-weights

    Opennessrestricted-weightsopen-weightscohere
  • Status changedModelGemini 3.1 Flash-Lite PreviewGoogle

    Gemini 3.1 Flash-Lite Preview: status changed from deprecated to retired

    Statusdeprecatedretiredgoogle
  • Openness changedModelLlama 3.1 8BMeta AI

    Llama 3.1 8B: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Openness changedModelLlama-3.2-1BMeta AI

    Llama-3.2-1B: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Openness changedModelLlama-3.1-405BMeta AI

    Llama-3.1-405B: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Openness changedModelLlama-3.2-3BMeta AI

    Llama-3.2-3B: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Openness changedModelLlama 4 ScoutMeta AI

    Llama 4 Scout: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Openness changedModelGemma 3 4BGoogle

    Gemma 3 4B: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Openness changedModeltiny-aya-globalCohere

    tiny-aya-global: openness changed from open-weights to restricted

    Opennessopen-weightsrestrictedhuggingface
  • Capabilities changedModelMistral Medium 3.1Mistral AI

    Mistral Medium 3.1: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-5 Image MiniOpenAI

    GPT-5 Image Mini: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelo3OpenAI

    o3: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-4.1-nanoOpenAI

    gpt-4.1-nano: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-5.6 Luna ProOpenAI

    GPT-5.6 Luna Pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelClaude Fable 5Anthropic

    Claude Fable 5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-5.6 Sol ProOpenAI

    GPT-5.6 Sol Pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT Chat LatestOpenAI

    GPT Chat Latest: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelFugu MaxSakana

    Fugu Max: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelMistral Medium 3Mistral AI

    Mistral Medium 3: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Context window changedModelMistral Small 4Mistral AI

    Mistral Small 4: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Capabilities changedModelgpt-5.5OpenAI

    gpt-5.5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGrok 4.3xAI

    Grok 4.3: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-4o-miniOpenAI

    gpt-4o-mini: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Context window changedModelgpt-5.6-terraOpenAI

    gpt-5.6-terra: context length changed from 1000000 to 1050000

    Context window1M tokens1.05M tokensopenrouter
  • Capabilities changedModelGPT-4o (2024-11-20)OpenAI

    GPT-4o (2024-11-20): modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelSakana NamazuSakana

    Sakana Namazu: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5-miniOpenAI

    gpt-5-mini: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGrok 4.6xAI

    Grok 4.6: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5OpenAI

    gpt-5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-4.1-miniOpenAI

    gpt-4.1-mini: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-6 Astra ProOpenAI

    GPT-6 Astra Pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-4o (2024-08-06)OpenAI

    GPT-4o (2024-08-06): modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5-nanoOpenAI

    gpt-5-nano: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.2-proOpenAI

    gpt-5.2-pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelo3-proOpenAI

    o3-pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.1OpenAI

    gpt-5.1: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.3-codexOpenAI

    gpt-5.3-codex: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelMuse Spark 1.3 ContributorMeta AI

    Muse Spark 1.3 Contributor: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelGrok 4.20xAI

    Grok 4.20: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Context window changedModelgpt-6-astraOpenAI

    gpt-6-astra: context length changed from 1000000 to 1050000

    Context window1M tokens1.05M tokensopenrouter
  • Capabilities changedModelGPT-5 ImageOpenAI

    GPT-5 Image: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGrok 4.20 Multi-AgentxAI

    Grok 4.20 Multi-Agent: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelClaude Opus 4.7Anthropic

    Claude Opus 4.7: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelVoxtral Small 24B 2507Mistral AI

    Voxtral Small 24B 2507: modalities changed from ["audio", "text"] to ["audio", "document", "text"]

    Modalitiesaudio, textaudio, document, textopenrouter
  • Capabilities changedModelClaude Opus 4.6Anthropic

    Claude Opus 4.6: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelMixtral 8x22B InstructMistral AI

    Mixtral 8x22B Instruct: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelClaude Opus 4Anthropic

    Claude Opus 4: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelo4-miniOpenAI

    o4-mini: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Context window changedModelZ.ai GLM 5.2Z.ai (Zhipu AI)

    Z.ai GLM 5.2: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelgpt-5.6-solOpenAI

    gpt-5.6-sol: context length changed from 1000000 to 1050000

    Context window1M tokens1.05M tokensopenrouter
  • Capabilities changedModelgpt-4.1OpenAI

    gpt-4.1: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-5.6 Terra ProOpenAI

    GPT-5.6 Terra Pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelMuse Spark 1.2Meta AI

    Muse Spark 1.2: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelClaude Opus 4.1Anthropic

    Claude Opus 4.1: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelNova 2 LiteAmazon Web Services

    Nova 2 Lite: modalities changed from ["image", "text", "video"] to ["document", "image", "text", "video"]

    Modalitiesimage, text, videodocument, image, text, videoopenrouter
  • Capabilities changedModelMuse Spark 1.2 ContributorMeta AI

    Muse Spark 1.2 Contributor: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelCodestral 2508Mistral AI

    Codestral 2508: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelMuse Spark 1.3Meta AI

    Muse Spark 1.3: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelMistral Large 3 2512Mistral AI

    Mistral Large 3 2512: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelClaude Opus 4.5Anthropic

    Claude Opus 4.5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGrok 4.5xAI

    Grok 4.5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelMuse Spark 1.1Meta AI

    Muse Spark 1.1: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelDevstral 2Mistral AI

    Devstral 2: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelMistral Large 2.0Mistral AI

    Mistral Large 2.0: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelGemini 2.5 Pro Preview 05-06Google

    Gemini 2.5 Pro Preview 05-06: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelo1-proOpenAI

    o1-pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-4o (2024-05-13)OpenAI

    GPT-4o (2024-05-13): modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGrok Build 0.1xAI

    Grok Build 0.1: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-4oOpenAI

    gpt-4o: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-4o-mini (2024-07-18)OpenAI

    GPT-4o-mini (2024-07-18): modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelMistral SabaMistral AI

    Mistral Saba: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelMistral LargeMistral AI

    Mistral Large: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelGPT-5.2 ChatOpenAI

    GPT-5.2 Chat: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelClaude Sonnet 4Anthropic

    Claude Sonnet 4: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelClaude Sonnet 4.6Anthropic

    Claude Sonnet 4.6: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5-proOpenAI

    gpt-5-pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.4OpenAI

    gpt-5.4: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelFugu Ultra v2Sakana

    Fugu Ultra v2: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Context window changedModelgpt-5.6-lunaOpenAI

    gpt-5.6-luna: context length changed from 1000000 to 1050000

    Context window1M tokens1.05M tokensopenrouter
  • Capabilities changedModelo1OpenAI

    o1: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.4-nanoOpenAI

    gpt-5.4-nano: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.4-miniOpenAI

    gpt-5.4-mini: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.2OpenAI

    gpt-5.2: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelgpt-5.4-proOpenAI

    gpt-5.4-pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelClaude Opus 4.8Anthropic

    Claude Opus 4.8: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGemini 3.1 Pro Preview Custom ToolsGoogle

    Gemini 3.1 Pro Preview Custom Tools: modalities changed from ["audio", "image", "text", "video"] to ["audio", "document", "image", "text", "video"]

    Modalitiesaudio, image, text, videoaudio, document, image, text, videoopenrouter
  • Capabilities changedModelMistral Medium 3.5Mistral AI

    Mistral Medium 3.5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGemini 2.5 Pro Preview 06-05Google

    Gemini 2.5 Pro Preview 06-05: modalities changed from ["audio", "image", "text"] to ["audio", "document", "image", "text"]

    Modalitiesaudio, image, textaudio, document, image, textopenrouter
  • Capabilities changedModelClaude Sonnet 4.5Anthropic

    Claude Sonnet 4.5: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelGPT-5.4 Image 2OpenAI

    GPT-5.4 Image 2: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Capabilities changedModelo3-miniOpenAI

    o3-mini: modalities changed from ["text"] to ["document", "text"]

    Modalitiestextdocument, textopenrouter
  • Capabilities changedModelgpt-5.5-proOpenAI

    gpt-5.5-pro: modalities changed from ["image", "text"] to ["document", "image", "text"]

    Modalitiesimage, textdocument, image, textopenrouter
  • Context window changedModelGemini 3.1 Flash-Lite PreviewGoogle

    Gemini 3.1 Flash-Lite Preview: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensgoogle
  • Context window changedModelGemini 2.5 Flash-LiteGoogle

    Gemini 2.5 Flash-Lite: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensgoogle
  • Context window changedModelGemini 2.5 FlashGoogle

    Gemini 2.5 Flash: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensgoogle
  • Context window changedModelGemini 3.5 FlashGoogle

    Gemini 3.5 Flash: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensgoogle
  • Context window changedModelGemini 3.7 FlashGoogle

    Gemini 3.7 Flash: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensgoogle
  • Context window changedModelGemini 3.8 FlashGoogle

    Gemini 3.8 Flash: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensgoogle
  • Context window changedModelQwen3 235B A22BQwen

    Qwen3 235B A22B: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Version releasedFrameworkphoenixArize AI

    phoenix: latest version changed from arize-phoenix: v20.10.0 to arize-phoenix: v20.11.0

    Latest versionarize-phoenix: v20.10.0arize-phoenix: v20.11.0github
  • Version releasedFrameworkpydantic-aiPydantic

    pydantic-ai: latest version changed from v2.42.0 (2026-09-08) to v2.43.0 (2026-09-11)

    Latest versionv2.42.0 (2026-09-08)v2.43.0 (2026-09-11)github
  • Version releasedFrameworkgemini-cliGoogle

    gemini-cli: latest version changed from Release v0.61.0-nightly.20260911.ged2ac40df to Release v0.61.0-nightly.20260912.g9c1b0a610

    Latest versionRelease v0.61.0-nightly.20260911.ged2ac40dfRelease v0.61.0-nightly.20260912.g9c1b0a610github
  • Version releasedFrameworkllama.cppggml.ai

    llama.cpp: latest version changed from b10917 to b10919

    Latest versionb10917b10919github
  • Version releasedFrameworkpytorchPyTorch Foundation

    pytorch: latest version changed from viable/strict/1789165518 to trunk/363267c09a72be003607d9e0e50df93150e4cb9a: [torchcom…

    Latest versionviable/strict/1789165518trunk/363267c09a72be003607d9e0e50df93150e4cb9a: [torchcomms hash update] update the pinned torchcommgithub
  • Context window changedModelDeepSeek V3.2DeepSeek

    DeepSeek V3.2: context length changed from 128000 to 163840

    Context window128K tokens163.8K tokensopenrouter
  • Context window changedModelHy3 previewTencent

    Hy3 preview: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelHermes 4 405BNous Research

    Hermes 4 405B: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelMiniMax-M3MiniMax

    MiniMax M3: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelMiMo-V2.5-ProXiaomi

    MiMo-V2.5-Pro: context length changed from 1000000 to 1050000

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelInklingThinking Machines

    Inkling: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelQwen3 MaxQwen

    Qwen3 Max Thinking: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelKimi K2.7 CodeMoonshot AI

    Kimi K2.7 Code: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelQwen3.6 Max PreviewQwen

    Qwen3.6 Max Preview: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelSolar Pro 3Upstage

    Solar Pro 3: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelHy3Tencent

    Hy3: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelMiMo-V2.5Xiaomi

    MiMo-V2.5: context length changed from 1000000 to 1050000

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelGLM 4.7Z.ai (Zhipu AI)

    GLM 4.7: context length changed from 200000 to 204800

    Context window200K tokens204.8K tokensopenrouter
  • Context window changedModelSolar Pro 4Upstage

    Solar Pro 4: context length changed from 512000 to 524288

    Context window512K tokens524.3K tokensopenrouter
  • Context window changedModelInkling SmallThinking Machines

    Inkling Small: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelQwen3.8 27BQwen

    Qwen3.8 27B: context length changed from 256000 to 1000000

    Context window256K tokens1M tokensopenrouter
  • Context window changedModelgpt-5.5-proOpenAI

    gpt-5.5-pro: context length changed from 922000 to 1050000

    Context window922K tokens1.05M tokensopenrouter
  • Context window changedModelQwen3.8 2.4T A95BQwen

    Qwen3.8 2.4T A95B: context length changed from 983616 to 1048576

    Context window983.6K tokens1.05M tokensopenrouter
  • Context window changedModelgpt-3.5-turboOpenAI

    GPT-3.5 Turbo: context length changed from 4096 to 16385

    Context window4.1K tokens16.4K tokensopenrouter
  • Context window changedModelGLM 5.3 FlashZ.ai (Zhipu AI)

    GLM 5.3 Flash: context length changed from 1000000 to 1310720

    Context window1M tokens1.31M tokensopenrouter
  • Context window changedModelgpt-5.5OpenAI

    gpt-5.5: context length changed from 922000 to 1050000

    Context window922K tokens1.05M tokensopenrouter
  • Context window changedModelLongCat 2.0Meituan

    LongCat 2.0: context length changed from 1000000 to 1048756

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelGLM 5.3Z.ai (Zhipu AI)

    GLM 5.3: context length changed from 1000000 to 1310720

    Context window1M tokens1.31M tokensopenrouter
  • Context window changedModelSonar Reasoning ProPerplexity AI

    Sonar Reasoning Pro: context length changed from 127000 to 128000

    Context window127K tokens128K tokensopenrouter
  • Context window changedModelNVIDIA Nemotron 3.5 Lightning 30B A3BNVIDIA

    Nemotron 3.5 Lightning: context length changed from 1000000 to 262144

    Context window1M tokens262.1K tokensopenrouter
  • Context window changedModelLing 3.0 Flash VLinclusionAI

    Ling 3.0 Flash VL: context length changed from 262144 to 131072

    Context window262.1K tokens131.1K tokensopenrouter
  • Context window changedModelMuse Spark 1.3Meta AI

    Muse Spark 1.3: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelGemma 3 12BGoogle

    Gemma 3 12B: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelReka Flash 3rekaai

    Reka Flash 3: context length changed from 128000 to 65536

    Context window128K tokens65.5K tokensopenrouter
  • Context window changedModelMistral LargeMistral AI

    Mistral Large: context length changed from 32768 to 128000

    Context window32.8K tokens128K tokensopenrouter
  • Context window changedModelNemotron 3 Nano 30B A3BNVIDIA

    Nemotron 3 Nano 30B A3B: context length changed from 1000000 to 262144

    Context window1M tokens262.1K tokensopenrouter
  • Context window changedModelDeepSeek V3.1DeepSeek

    DeepSeek V3.1: context length changed from 128000 to 163840

    Context window128K tokens163.8K tokensopenrouter
  • Context window changedModelGemma 3 27BGoogle

    Gemma 3 27B: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelQwen3.8 FlashQwen

    Qwen3.8 Flash: context length changed from 256000 to 1000000

    Context window256K tokens1M tokensopenrouter
  • Context window changedModelDeepSeek V3 0324DeepSeek

    DeepSeek V3 0324: context length changed from 128000 to 163840

    Context window128K tokens163.8K tokensopenrouter
  • Context window changedModelGemma 3 4BGoogle

    Gemma 3 4B: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelMistral SabaMistral AI

    Mistral Saba: context length changed from 32000 to 32768

    Context window32K tokens32.8K tokensopenrouter
  • Context window changedModelQwen3 32BQwen

    Qwen3 32B: context length changed from 32768 to 131072

    Context window32.8K tokens131.1K tokensopenrouter
  • Context window changedModelgpt-4.1OpenAI

    gpt-4.1: context length changed from 1000000 to 1047576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelMistral Large 2.0Mistral AI

    Mistral Large 2.0: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelGLM 4.5Z.ai (Zhipu AI)

    GLM 4.5: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelGLM 4.5 AirZ.ai (Zhipu AI)

    GLM 4.5 Air: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelgpt-4.1-miniOpenAI

    gpt-4.1-mini: context length changed from 1000000 to 1047576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelQwen2.5 Coder 32B InstructQwen

    Qwen2.5 Coder 32B Instruct: context length changed from 131072 to 32768

    Context window131.1K tokens32.8K tokensopenrouter
  • Context window changedModelKimi K2 0905Moonshot AI

    Kimi K2 0905: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelGLM 4.5VZ.ai (Zhipu AI)

    GLM 4.5V: context length changed from 64000 to 65536

    Context window64K tokens65.5K tokensopenrouter
  • Context window changedModelgpt-4.1-nanoOpenAI

    gpt-4.1-nano: context length changed from 1000000 to 1047576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelGLM 4.6Z.ai (Zhipu AI)

    GLM 4.6: context length changed from 200000 to 204800

    Context window200K tokens204.8K tokensopenrouter
  • Context window changedModelQwen3 235B A22B Instruct 2507Qwen

    Qwen3 235B A22B Instruct 2507: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelDeepSeek V3DeepSeek

    DeepSeek V3: context length changed from 128000 to 163840

    Context window128K tokens163.8K tokensopenrouter
  • Context window changedModelHermes 3 70B InstructNous Research

    Hermes 3 70B Instruct: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelGLM 5Z.ai (Zhipu AI)

    GLM 5: context length changed from 200000 to 204800

    Context window200K tokens204.8K tokensopenrouter
  • Context window changedModelLlama 4 MaverickMeta AI

    Llama 4 Maverick: context length changed from 1000000 to 1048576

    Context window1M tokens1.05M tokensopenrouter
  • Context window changedModelKimi K2 0711Moonshot AI

    Kimi K2 0711: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelR1 Distill Llama 70BDeepSeek

    R1 Distill Llama 70B: context length changed from 128000 to 8192

    Context window128K tokens8.19K tokensopenrouter
  • Context window changedModelQwen3 Coder NextQwen

    Qwen3 Coder Next: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelStep 3.5 FlashStepFun

    Step 3.5 Flash: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelGranite 4.0 MicroIBM

    Granite 4.0 Micro: context length changed from 128000 to 131000

    Context window128K tokens131K tokensopenrouter
  • Context window changedModelQwen3 VL 8B InstructQwen

    Qwen3 VL 8B Instruct: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelDeepSeek V3.1 TerminusDeepSeek

    DeepSeek V3.1 Terminus: context length changed from 128000 to 163840

    Context window128K tokens163.8K tokensopenrouter
  • Context window changedModelMistral Medium 3Mistral AI

    Mistral Medium 3: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelPhi 4Microsoft

    Phi 4: context length changed from 16000 to 16384

    Context window16K tokens16.4K tokensopenrouter
  • Context window changedModelgpt-3.5-turboOpenAI

    gpt-3.5-turbo: context length changed from 4096 to 4095

    Context window4.1K tokens4.1K tokensopenrouter
  • Context window changedModelGemma 4 31BGoogle

    Gemma 4 31B: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelGLM 5V TurboZ.ai (Zhipu AI)

    GLM 5V Turbo: context length changed from 200000 to 202752

    Context window200K tokens202.8K tokensopenrouter
  • Context window changedModelTrinity Large ThinkingArcee AI

    Trinity Large Thinking: context length changed from 512000 to 262144

    Context window512K tokens262.1K tokensopenrouter
  • Context window changedModelGLM 5 TurboZ.ai (Zhipu AI)

    GLM 5 Turbo: context length changed from 200000 to 202752

    Context window200K tokens202.8K tokensopenrouter
  • Context window changedModelGemma 4 26B A4BGoogle

    Gemma 4 26B A4B: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelKAT-Coder-Pro V2Kwaipilot

    KAT-Coder-Pro V2: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelNemotron 3 SuperNVIDIA

    Nemotron 3 Super: context length changed from 1000000 to 262144

    Context window1M tokens262.1K tokensopenrouter
  • Context window changedModelLlama 4 ScoutMeta AI

    Llama 4 Scout: context length changed from 10000000 to 1310720

    Context window10M tokens1.31M tokensopenrouter
  • Context window changedModelGLM 5.1Z.ai (Zhipu AI)

    GLM 5.1: context length changed from 200000 to 204800

    Context window200K tokens204.8K tokensopenrouter
  • Context window changedModelKimi K2.5Moonshot AI

    Kimi K2.5: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelGLM 4.6VZ.ai (Zhipu AI)

    GLM 4.6V: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelQwen3 VL 30B A3B InstructQwen

    Qwen3 VL 30B A3B Instruct: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelgpt-5.1OpenAI

    gpt-5.1: context length changed from 272000 to 400000

    Context window272K tokens400K tokensopenrouter
  • Context window changedModelDevstral 2Mistral AI

    Devstral 2: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelDeepSeek V3.2 ExpDeepSeek

    DeepSeek V3.2 Exp: context length changed from 128000 to 163840

    Context window128K tokens163.8K tokensopenrouter
  • Context window changedModelKimi K2 ThinkingMoonshot AI

    Kimi K2 Thinking: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Context window changedModelMistral Small 3Mistral AI

    Mistral Small 3: context length changed from 32000 to 32768

    Context window32K tokens32.8K tokensopenrouter
  • Context window changedModelMistral Medium 3.1Mistral AI

    Mistral Medium 3.1: context length changed from 128000 to 131072

    Context window128K tokens131.1K tokensopenrouter
  • Context window changedModelSonarPerplexity AI

    Sonar: context length changed from 127000 to 127072

    Context window127K tokens127.1K tokensopenrouter
  • Context window changedModelQwen3 14BQwen

    Qwen3 14B: context length changed from 32768 to 131072

    Context window32.8K tokens131.1K tokensopenrouter
  • Context window changedModelgpt-4OpenAI

    gpt-4: context length changed from 8192 to 8191

    Context window8.19K tokens8.19K tokensopenrouter
  • Context window changedModelQwen3 VL 32B InstructQwen

    Qwen3 VL 32B Instruct: context length changed from 256000 to 131072

    Context window256K tokens131.1K tokensopenrouter
  • Context window changedModelKimi K2.6Moonshot AI

    Kimi K2.6: context length changed from 256000 to 262144

    Context window256K tokens262.1K tokensopenrouter
  • Version releasedFrameworkpytorchPyTorch Foundation

    pytorch: latest version changed from viable/strict/1789158535: [ROCm][CD] Add gfx1250 (MI450) … to viable/strict/1789165518

    Latest versionviable/strict/1789158535: [ROCm][CD] Add gfx1250 (MI450) to the nightly wheel arch list (#196610)viable/strict/1789165518github
  • Context window changedModelQwen2.5 72B InstructQwen

    Qwen2.5 72B Instruct: context length changed from 131072 to 32768

    Context window131.1K tokens32.8K tokensopenrouter
  • Capabilities changedModelGemini 3 Pro PreviewGoogle

    Gemini 3 Pro Preview: capabilities changed from ["Caching", "Code execution", "File search", "Function ca… to ["caching", "code_execution", "file_search", "function_ca…

    CapabilitiesCaching, Code execution, File search, Function calling, Search grounding, Structured outputs, Thinking, URL contextcaching, code_execution, file_search, function_calling, reasoning, search_grounding, structured_output, url_contextgoogle
  • Capabilities changedModelGemini 3.1 Flash-Lite PreviewGoogle

    Gemini 3.1 Flash-Lite Preview: capabilities changed from ["Caching", "Code execution", "File search", "Function ca… to ["caching", "code_execution", "file_search", "function_ca…

    CapabilitiesCaching, Code execution, File search, Function calling, Grounding with Google Maps, Search grounding, Structured outputs, Thinking, URL contextcaching, code_execution, file_search, function_calling, maps_grounding, reasoning, search_grounding, structured_output, url_contextgoogle
  • Capabilities changedModelGemini 2.0 Flash-LiteGoogle

    Gemini 2.0 Flash-Lite: capabilities changed from ["Caching", "Function calling", "Structured outputs"] to ["caching", "function_calling", "structured_output"]

    CapabilitiesCaching, Function calling, Structured outputscaching, function_calling, structured_outputgoogle
  • Capabilities changedModelGemini 2.0 FlashGoogle

    Gemini 2.0 Flash: capabilities changed from ["Caching", "Code execution", "Function calling", "Ground… to ["caching", "code_execution", "function_calling", "maps_g…

    CapabilitiesCaching, Code execution, Function calling, Grounding with Google Maps, Search grounding, Structured outputscaching, code_execution, function_calling, maps_grounding, search_grounding, structured_outputgoogle
  • Capabilities changedModelGemini Robotics ER 1.6Google

    Gemini Robotics ER 1.6: capabilities changed from ["Caching", "Code execution", "Computer use", "File searc… to ["caching", "code_execution", "computer_use", "file_searc…

    CapabilitiesCaching, Code execution, Computer use, File search, Function calling, Grounding with Google Maps, Search grounding, Structured outputs, Thinking, URL contextcaching, code_execution, computer_use, file_search, function_calling, maps_grounding, reasoning, search_grounding, structured_output, url_contextgoogle
  • Capabilities changedModelGemini Robotics ER 2 PreviewGoogle

    Gemini Robotics ER 2 Preview: capabilities changed from ["Caching", "Code execution", "Computer use", "File searc… to ["caching", "code_execution", "computer_use", "file_searc…

    CapabilitiesCaching, Code execution, Computer use, File search, Function calling, Grounding with Google Maps, Search grounding, Structured outputs, Thinking, URL contextcaching, code_execution, computer_use, file_search, function_calling, maps_grounding, reasoning, search_grounding, structured_output, url_contextgoogle
  • Capabilities changedModelLyria 3Google

    Lyria 3: capabilities changed from ["Audio generation"] to ["audio_output"]

    CapabilitiesAudio generationaudio_outputgoogle
  • Capabilities changedModelLyria 3.5Google

    Lyria 3.5: capabilities changed from ["Audio generation"] to ["audio_output"]

    CapabilitiesAudio generationaudio_outputgoogle

Other new entities

Other new entities 107

Other new entities
EntityTypeOrganizationFirst seen
Llama 3.1FamilySao10K12 Sept 2026
Llama 3.3FamilySao10K12 Sept 2026
SeedFamilyNaver12 Sept 2026
LlamaFamilyDeepSeek12 Sept 2026
Qwen3FamilyDeepSeek12 Sept 2026
Llama 3.1FamilyMeta AI12 Sept 2026
Llama 3.3FamilyMeta AI12 Sept 2026
MistralFamilyNous Research12 Sept 2026
Llama 3.1FamilyNVIDIA12 Sept 2026
QwenFamilyQwen12 Sept 2026
Claude 2.0FamilyAnthropic12 Sept 2026
LlamaFamilyNVIDIA12 Sept 2026
Claude 4.1FamilyAnthropic12 Sept 2026
BigScience BLOOM RAIL 1.0License12 Sept 2026
BSD 3-Clause LicenseLicense12 Sept 2026
Apache License 2.0License12 Sept 2026
Other (unclassified licence)License12 Sept 2026
GNU GPL v3License12 Sept 2026
Gemma Terms of UseLicense12 Sept 2026
BSD 2-Clause LicenseLicense12 Sept 2026
Llama 3.2 Community LicenseLicense12 Sept 2026
Modified MIT LicenseLicense12 Sept 2026
MIT LicenseLicense12 Sept 2026
GNU AGPL v3License12 Sept 2026
Creative Commons Attribution-NonCommercial 4.0License12 Sept 2026
Creative Commons Attribution 4.0License12 Sept 2026
Llama 3.3 Community LicenseLicense12 Sept 2026
Llama 3.1 Community LicenseLicense12 Sept 2026
Apple Sample Code / ML Research LicenseLicense12 Sept 2026
Llama 3 Community LicenseLicense12 Sept 2026
Llama 2 Community LicenseLicense12 Sept 2026
Creative Commons Attribution-NonCommercial-ShareAlike 4.0License12 Sept 2026
Open RAIL++-M (Stable Diffusion XL)License12 Sept 2026
Creative Commons Attribution-ShareAlike 4.0License12 Sept 2026
Proprietary (API / terms of service)License12 Sept 2026
Research / non-commercial licence (custom)License12 Sept 2026
SWE-Bench ProBenchmark12 Sept 2026
MMTEBBenchmark12 Sept 2026
Terminal-Bench 2.0Benchmark12 Sept 2026
ARC-AGI-2Benchmark12 Sept 2026
AIME 2024Benchmark12 Sept 2026
LiveBench Instruction FollowingBenchmark12 Sept 2026
LiveBench LanguageBenchmark12 Sept 2026
LiveBench Data AnalysisBenchmark12 Sept 2026
LiveBench MathematicsBenchmark12 Sept 2026
LiveBench Agentic CodingBenchmark12 Sept 2026
LiveBench CodingBenchmark12 Sept 2026
LiveBench ReasoningBenchmark12 Sept 2026
Aider polyglot — well-formed responsesBenchmark12 Sept 2026
GPQA DiamondBenchmark12 Sept 2026
KuaishouCompany12 Sept 2026
Qwen APIProvider12 Sept 2026
Mistral AIProvider12 Sept 2026
Skild AICompany12 Sept 2026
Vittoria CavicchioliResearcher12 Sept 2026
Michele PestarinoResearcher12 Sept 2026
Davide MalvezziResearcher12 Sept 2026
Matt WhiteResearcher12 Sept 2026
Yiyun SuResearcher12 Sept 2026
Zelin LiResearcher12 Sept 2026
Daniel KirchnerResearcher12 Sept 2026
Christoph BenzmuellerResearcher12 Sept 2026
Christopher BuccafuscoResearcher12 Sept 2026
Emily WengerResearcher12 Sept 2026
Nirav PatelResearcher12 Sept 2026
Paul LernerResearcher12 Sept 2026
Salim HafidResearcher12 Sept 2026
Pierre-Antoine LequeuResearcher12 Sept 2026
Hucheng YangResearcher12 Sept 2026
Lang AnResearcher12 Sept 2026
Yixiong XiaoResearcher12 Sept 2026
Hang LyuResearcher12 Sept 2026
Matteo VaccargiuResearcher12 Sept 2026
Daniel GraziotinResearcher12 Sept 2026
Giuseppe DestefanisResearcher12 Sept 2026
Cato Elia KurtzResearcher12 Sept 2026
Blake G. FitchResearcher12 Sept 2026
Weiran WangResearcher12 Sept 2026
Zhouyuan HuoResearcher12 Sept 2026
Aaron Isidore GraceResearcher12 Sept 2026
Zhangkai WuResearcher12 Sept 2026
Jiaqi ZhangResearcher12 Sept 2026
Yiqi WangResearcher12 Sept 2026
Benedikt BolligResearcher12 Sept 2026
Ronghe WangResearcher12 Sept 2026
Priyam SrivastavaResearcher12 Sept 2026
Xin JinResearcher12 Sept 2026
Yuhang GuResearcher12 Sept 2026
Chengtao LaiResearcher12 Sept 2026
Zhongchun ZhouResearcher12 Sept 2026
Beatriz Navarro LamedaResearcher12 Sept 2026
Nikoleta KalaydzhievaResearcher12 Sept 2026
Benjamin J. WalkerResearcher12 Sept 2026
Sebastian SzyllerResearcher12 Sept 2026
Vasisht DudduResearcher12 Sept 2026
Asim WaheedResearcher12 Sept 2026
Hridoy Sankar DuttaResearcher12 Sept 2026
Keshav SoodResearcher12 Sept 2026
Kamel KamelResearcher12 Sept 2026
Yinghao TangResearcher12 Sept 2026
Luoxuan WengResearcher12 Sept 2026
Zhen WenResearcher12 Sept 2026
Tommy ShaResearcher12 Sept 2026
Stella ZhaoResearcher12 Sept 2026
Baokun WangResearcher12 Sept 2026
Jinyong WenResearcher12 Sept 2026
Jinsong ShuResearcher12 Sept 2026

Gone entities

Gone entities 1

“Gone” means an entity present at the first date is no longer current at the second (merged or retired) — its record and history are kept. Events are keyed on occurred_at; back-filled history is excluded (tick “include backfill” to add it). Per-entity history: any entity's History tab. Atlas as of 5 Sept 2026 · Methodology →