Skip to main content

HomeSubscribe vs API

Explore models

Decision surfaces

Models workbenchPrice, performance and catalog filters Model catalogSearch, filter and compare model evidence Lifecycle radarRetirements, sunset dates and migration paths

Top 1-5

Live rank · unavailable

Ranked models are unavailable.

Top 6-10

Live rank · unavailable

Ranked models are unavailable.

Featured

Start here

All leaderboardsEvery ranking on ModelScale, by capability and by benchmark.Popular modelsEvery scored model, sortable across all seven capability axes, with quality-against-cost plots.Make it yoursWeight the axes and the runtime thresholds yourself, and rank the catalog by your own definition of best.

Capabilities

Ranked by axis

AgenticRanked for autonomous, multi-step task performance CodingRanked for software engineering task performance KnowledgeRanked for factual recall and domain breadth ReasoningRanked for multi-step logical reasoning

Benchmarks

Per-benchmark rankings

All benchmarks →

Compare

Pick a surface

Compare by ModelTwo to four models side by side: capability scores, runtime, every published price component and the source's own pair deltas.Compare by ProviderEvery model from two to four providers, ranked together on one capability table and one quality-versus-cost plot.

Featured model comparisons

Ready-made

Compare models →

Featured provider comparisons

Ready-made

Compare providers →

Articles

Everything about AI models

AllGuidesInsightsNews

LanguageTranslate this page

Search languages

Preferred languages

More languages supported by Google Translate

Open the selector to load the current language catalog.

Translations are provided by Google Translate.

Loading page…

Source-aware model, pricing, and workload evidence for practical AI decisions.

Explore

Models workbench Compare Lifecycle radar Subscribe vs API

Leaderboards

All leaderboards Popular models Make it yours Agentic Coding Knowledge Reasoning

Benchmarks

Resources

Articles Sources llms.txt

Elsewhere

X LinkedIn GitHub

Evidence values belong to their cited sources.Powered by MonoMind AI Lab

HomeSubscribe vs API

Explore models

Decision surfaces

Models workbenchPrice, performance and catalog filters Model catalogSearch, filter and compare model evidence Lifecycle radarRetirements, sunset dates and migration paths

Top 1-5

Live rank · 2026-10-23

#1Claude Opus 5.5Anthropic86.7 #2GPT-6.1 SolOpenAI84.2 #3GPT-6 AstraOpenAI84.0 #4Claude Fable 5.1Anthropic83.2 #5Gemini 4 ArgonGoogle82.9

Top 6-10

Live rank · 2026-10-23

#6Claude Sonnet 5.5Anthropic82.2 #7Claude Fable 5Anthropic81.3 #8Claude Opus 5Anthropic80.1 #9GPT-6 SolOpenAI78.6 #10GPT-5.6 SolOpenAI75.3

Featured

Start here

All leaderboardsEvery ranking on ModelScale, by capability and by benchmark.Popular modelsEvery scored model, sortable across all seven capability axes, with quality-against-cost plots.Make it yoursWeight the axes and the runtime thresholds yourself, and rank the catalog by your own definition of best.

Capabilities

Ranked by axis

AgenticRanked for autonomous, multi-step task performance CodingRanked for software engineering task performance KnowledgeRanked for factual recall and domain breadth ReasoningRanked for multi-step logical reasoning

Benchmarks

Per-benchmark rankings

SWE-bench Verified leaderboardSWE-bench Verified tests patches for 500 human-filtered GitHub issues.Terminal-Bench 2.0 leaderboardA benchmark for agentic software engineering tasks executed in real terminal environments.GPQA-D leaderboardA display-only GPQA Diamond reference from provider comparison charts.FrontierMath v2 (Tiers 1-3) leaderboardEpoch AI's corrected v2 core FrontierMath suite of private advanced mathematics problems.MMMU-Pro leaderboardA harder multimodal benchmark for frontier models that combines text with images, diagrams, charts, and academic visual reasoning tasks.ARC-AGI-2 leaderboardARC-AGI-2 tests whether a system can infer a transformation from example grids and apply it to a new grid.All benchmarks →

Compare

Pick a surface

Compare by ModelTwo to four models side by side: capability scores, runtime, every published price component and the source's own pair deltas.Compare by ProviderEvery model from two to four providers, ranked together on one capability table and one quality-versus-cost plot.

Featured model comparisons

Ready-made

Claude Opus 5.5 vs GPT-6.1 SolThe catalog's two highest-ranked models, head to head on every capability axis.GPT-6 Astra vs Claude Fable 5.1The next two models in the published ranking from different providers, one rung below the leaders.Qwen3.7 Flash vs GPT-OSS 20BThe two lowest-priced models holding a place on the published price-performance frontier.Claude Opus 5.5 vs MiMo-V2.6-ProThe highest-ranked open-weight model against the highest-ranked proprietary one.Claude Opus 5.5 vs GPT-OSS 20BThe top of the published ranking against the cheapest model still on the price-performance frontier.

Featured provider comparisons

Ready-made

OpenAI vs AnthropicThe two providers holding the highest-scoring models in the catalog, across their whole published ranges.Google vs MetaThe next two providers by their best published score, one tier below the leaders.OpenAI vs AlibabaThe two largest scored ranges in the catalog, every model of each on one table.Anthropic vs XiaomiThe highest-scoring open-weight provider against the highest-scoring proprietary one.Alibaba vs Xiaomi vs StepFunThe three highest-scoring open-weight providers at once, on one capability table and one plot.

Articles

Everything about AI models

AllGuidesInsightsNews

LanguageTranslate this page

Search languages

Preferred languages

More languages supported by Google Translate

Open the selector to load the current language catalog.

Translations are provided by Google Translate.

Lifecycle

Production model lifecycle and retirement radar

Review announced and observed model changes against their source evidence before choosing a migration path. Every date shown is the date a source published, not an estimate.

Retirement notices

19 of 19 notices in the selected horizon. Horizons are computed from the effective date the lifecycle source returned; a notice with no effective date never enters a dated horizon.

confirmed

claude-1.0 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-1.1 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-1.2 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-1.3 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-instant-1.0 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-instant-1.1 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-instant-1.2 retired

Anthropic

2024-11-06722 days ago

Announced 2024-09-04Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-2.0 retired

Anthropic

2025-07-21465 days ago

Announced 2025-01-21Replacement: Claude Opus 4.8 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-2.1 retired

Anthropic

2025-07-21465 days ago

Announced 2025-01-21Replacement: Claude Opus 4.8 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-sonnet-20240229 retired

Anthropic

2025-07-21465 days ago

Announced 2025-01-21Replacement: Claude Sonnet 5.5 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-5-sonnet-20240620 retired

Anthropic

2025-10-28366 days ago

Announced 2025-08-13Replacement: Claude Sonnet 5.5 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-5-sonnet-20241022 retired

Anthropic

2025-10-28366 days ago

Announced 2025-08-13Replacement: Claude Sonnet 5.5 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-opus-20240229 retired

Anthropic

2026-01-05297 days ago

Announced 2025-06-30Replacement: Claude Opus 4.8 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-5-haiku-20241022 retired

Anthropic

2026-02-19252 days ago

Announced 2025-12-19Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-7-sonnet-20250219 retired

Anthropic

2026-02-19252 days ago

Announced 2025-10-28Replacement: Claude Sonnet 5.5 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-3-haiku-20240307 retired

Anthropic

2026-04-20192 days ago

Announced 2026-02-19Replacement id claude-haiku-4-5-20251001 is outside the current catalog.Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-opus-4-20250514 retired

Anthropic

2026-06-15136 days ago

Announced 2026-04-14Replacement: Claude Opus 4.8 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-sonnet-4-20250514 retired

Anthropic

2026-06-15136 days ago

Announced 2026-04-14Replacement: Claude Sonnet 5.5 Anthropic model deprecationsObserved Unavailable · not-collected

confirmed

claude-opus-4-1-20250805 retired

Anthropic

2026-08-0585 days ago

Announced 2026-06-05Replacement: Claude Opus 4.8 Anthropic model deprecationsObserved Unavailable · not-collected

Release timeline

Dated release events, kept separate from retirement notices.

2026-10-17DeepSeek V4.1 Flash 1017DeepSeek · confirmed

2026-10-16Beam 1.0 501B-A23B open weightsReflection AI · confirmed

2026-10-15Nemotron 3.5 Super VL 120B A12BNVIDIA · confirmed

2026-10-15d1-ultraLiquid AI · confirmed

2026-10-14Ultramarine 1 FlashPaper Instruments · confirmed

2026-10-08Celeris-1 DecisionCeleris · confirmed

2026-10-07Claude Haiku 5.5Anthropic · confirmed

2026-10-01Pareto 26.10 PreviewUnbiased · confirmed

2026-10-01Perplexity Decider v1 27BPerplexity · confirmed

2026-09-30Gemini 4 ArgonGoogle · confirmed

2026-09-30GLiDEFastino · confirmed

2026-09-30Embed 5 Pro and FastCohere · confirmed

2026-09-29GPT-6.1 SolOpenAI · confirmed

2026-09-29Ling 3.1 FlashInclusionAI · confirmed

2026-09-28Claude Sonnet 5.5Anthropic · confirmed

2026-09-23Gemini 3.8 Flash TTS and Flash-Lite TTSGoogle · confirmed

2026-09-22GPT-6 Sol and GPT-6 LunaOpenAI · confirmed

2026-09-22Claude Opus 5.5Anthropic · confirmed

2026-09-21Grok 4.7xAI · confirmed

2026-09-20Step 5 PreviewStepFun · confirmed

2026-09-15Gemini 3.8 Live and Gemini 3.8 Live Extended ThinkingGoogle · confirmed

2026-09-10GPT-Live-1 in the APIOpenAI · confirmed

2026-09-10DeepSeek V4.1 FlashDeepSeek · confirmed

2026-09-03GPT-6 AstraOpenAI · confirmed

2026-09-02Quasar 438BMultiverse Computing · confirmed

2026-09-02Muse Spark 1.3Meta · confirmed

2026-09-02Gemini 3.8 Flash and Gemini 3.8 Flash CyberGoogle · confirmed

2026-09-01Claude Fable 5.1 and Claude Mythos 5.1Anthropic · confirmed

2026-08-26Gemini 3.5 TranscribeGoogle · confirmed

2026-08-13Toast 1Mixedbread · confirmed

2026-08-13Gemini 3.7 FlashGoogle · confirmed

2026-08-13DeepSeek V4 Pro 0813DeepSeek · confirmed

2026-07-31DeepSeek V4 Flash 0731DeepSeek · confirmed

2026-07-28GPT TranscribeOpenAI · confirmed

2026-07-28GPT Live TranscribeOpenAI · confirmed

2026-07-24Claude Opus 5Anthropic · confirmed

2026-07-22Celeris-1Celeris · confirmed

2026-07-21Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash CyberGoogle · confirmed

2026-07-09GPT-5.6OpenAI · confirmed

2026-07-06GPT Realtime 2.1OpenAI · confirmed

2026-07-06GPT Realtime 2.1 MiniOpenAI · confirmed

2026-06-30Claude Sonnet 5Anthropic · confirmed

2026-06-09Claude Fable 5 and Claude Mythos 5Anthropic · confirmed

2026-05-28Claude Opus 4.8Anthropic · confirmed

2026-04-23GPT-5.5 and GPT-5.5 ProOpenAI · confirmed

2026-04-23GPT Audio 1.5OpenAI · confirmed

2026-04-16Claude Opus 4.7Anthropic · confirmed

2026-04-15Gemini 3.1 Flash TTS PreviewGoogle · confirmed

2026-04-02Gemma 4Google · confirmed

2026-03-17GPT-5.4 mini and nanoOpenAI · confirmed

2026-03-05GPT-5.4 and GPT-5.4 ProOpenAI · confirmed

2026-03-03GPT-5.3 InstantOpenAI · confirmed

2026-02-19Gemini 3.1 ProGoogle · confirmed

2026-02-05GPT-5.3-CodexOpenAI · confirmed

2026-02-01Claude 4.6Anthropic · confirmed

Evidence boundary

Only lifecycle records returned by the API appear here. A missing notice does not prove a model has no retirement plan — it proves only that this source published none. Replacement links resolve to a profile only when the returned replacement id exists in the current catalog; otherwise the raw id is shown so it can still be traced. Horizons are derived from the effective date, never from an assumption about it.

Dataset generated

2026-10-23 13:17 UTC

Source-aware model, pricing, and workload evidence for practical AI decisions.

Explore

Models workbench Compare Lifecycle radar Subscribe vs API

Leaderboards

All leaderboards Popular models Make it yours Agentic Coding Knowledge Reasoning

Benchmarks

SWE-bench Verified leaderboard Terminal-Bench 2.0 leaderboard GPQA-D leaderboard FrontierMath v2 (Tiers 1-3) leaderboard MMMU-Pro leaderboard ARC-AGI-2 leaderboard

Resources

Articles Sources llms.txt

Elsewhere

X LinkedIn GitHub

Evidence values belong to their cited sources.Powered by MonoMind AI Lab