Claude Fable 5.2
Newanthropic/claude-fable-5.2-20261109
This model accepts text, image, and file inputs and provides text outputs. It has a context length of 1,000,000 tokens and supports tools and reasoning. Pricing is $8 per million input tokens and $40 per million output tokens.
- tool use
- structured outputs
- reasoning
At a glance
Specifications
What the API gives you
- Context window
- 1M tokens
- Max output
- 128K tokens
- Input price
- $8 / 1M tokens
- Output price
- $40 / 1M tokens
- Cache read
- $0.20 / 1M tokens
- Modalities
- text, image, file → text
- Knowledge cutoff
- —
- Released
- Nov 9, 2026
Measured evidence
Public benchmark record
- Design Arena Coding
- 1375 Elo2026-11-15
- Design Arena Full-stack Agent
- 1400 Elo2026-11-15
- Design Arena UI
- 1382 Elo2026-11-15
- Design Arena Website
- 1352 Elo2026-11-15
- FrontierMath Tiers 1-3 (Epoch AI)‡
- 95.1% ± 1.32026-11-06max effort, evaluated 2026-11-06
- GPQA Diamond (Epoch AI)‡
- 93.9% ± 1.62026-11-06max effort, evaluated 2026-11-06
- LMArena Agent‡
- 30.42026-11-11max effort
- LMArena Agent Steerability‡
- 332026-11-11max effort
- LMArena Agent Task Outcome‡
- 222026-11-11max effort
- LMArena Agent Tool Hallucination‡
- 12026-11-11max effort
- LMArena WebDev‡
- 1838 Elo2026-11-11xhigh effort
- OTIS Mock AIME (Epoch AI)‡
- 99.7% ± 0.32026-11-06max effort, evaluated 2026-11-06
- SimpleQA Verified (Epoch AI)‡
- 94% ± 0.82026-11-06max effort, evaluated 2026-11-06
- Terminal-Bench (4.1)‡
- 65.8%2026-11-15Claude Code, max effort, evaluated 2026-11-13
LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open. Design Arena uses blind human preference. ‡ Submitted scaffold plus model result. Compare the harness and effort settings, not only the model name.
Task evidence
How broad is the support?
Each status is exact to this model version and the selected task. Official release claims remain provisional.
- Best LLM for Frontend & UIEstablished evidence
82% of intended weighted task evidence across 2 independent sources. 6 of 6 configured evaluation feeds currently have results.
- Best LLM for Autonomous AgentsEstablished evidence
54% of intended weighted task evidence across 2 independent sources. 6 of 6 configured evaluation feeds currently have results.
- Best LLM for Hermes AgentEstablished evidence
48% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.
- Best LLM for OpenHandsEstablished evidence
48% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.
- Best LLM for GooseEstablished evidence
48% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.
- Best LLM for OpenClawEstablished evidence
48% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.
Show all 21 tasks
- Best LLM for CodingEstablished evidence
47% of intended weighted task evidence across 3 independent sources. 7 of 7 configured evaluation feeds currently have results.
- Best LLM for Agentic CodingEstablished evidence
46% of intended weighted task evidence across 3 independent sources. 7 of 7 configured evaluation feeds currently have results.
- Best LLM for AI AgentsEstablished evidence
41% of intended weighted task evidence across 2 independent sources. 9 of 9 configured evaluation feeds currently have results.
- Best LLM for Competition MathEstablished evidence
30% of intended weighted task evidence across 2 independent sources. 6 of 6 configured evaluation feeds currently have results.
- Best LLM for Math & ReasoningEstablished evidence
24% of intended weighted task evidence across 3 independent sources. 7 of 7 configured evaluation feeds currently have results.
- Best LLM for Tool & Function CallingProvisional evidence
38% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Search. 6 of 7 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
- Best LLM for Code CompletionProvisional evidence
36% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 3 of 3 configured evaluation feeds currently have results.
- Best Local LLM for CodingProvisional evidence
21% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 4 of 4 configured evaluation feeds currently have results.
- Best LLM for Logic & Problem SolvingProvisional evidence
18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
- Best LLM for Debugging & Code ReviewProvisional evidence
17% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Computer & Browser UseProvisional evidence
17% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or OSWorld Verified. 3 of 3 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
- Best LLM for Grounded RAGProvisional evidence
14% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from CRAG or FACTS Grounding. 5 of 7 configured evaluation feeds currently have results.
- Best LLM for a General AssistantProvisional evidence
9% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Reasoning. 6 of 6 configured evaluation feeds currently have results.
- Best LLM for Structured Data ExtractionProvisional evidence
Official release evidence found; independent confirmation is still missing from Berkeley Function Calling or LiveBench Data Analysis.
- Best LLM for Charts & DiagramsProvisional evidence
Official release evidence found; independent confirmation is still missing from LMArena Document or LMArena Vision.
Official release evidence · provisional
What the launch material actually says
Verbatim claims from the model provider. These help on launch day but do not count as independent confirmation.
“Claude Fable 5.2 is the strongest model we have measured on data-eng-bench, our open-source benchmark of 103 data-engineering tasks.”
Anthropic · Introducing Claude Fable 5.2 and Claude Mythos 5.2 \ Anthropic“Fable 5.2 Fable 5.1 Opus 5.5 GPT-6 Astra Gemini 4 Carbon Agentic coding FrontierCode 2.0 Agentic coding FrontierCode 2.0 49.1% Xhigh effort 32.4% Xhigh effort 45.5% 42.8% 41.1%”
Anthropic · Introducing Claude Fable 5.2 and Claude Mythos 5.2 \ Anthropic“Agentic coding FrontierSWE v2 Agentic coding FrontierSWE v2 69.9% 56.3% 62.3% 65.5% 61.2%”
Anthropic · Introducing Claude Fable 5.2 and Claude Mythos 5.2 \ Anthropic“Business workflows AutomationBench Business workflows AutomationBench 51.4% 31.4% 42.5% 41.4% 53.9%”
Anthropic · Introducing Claude Fable 5.2 and Claude Mythos 5.2 \ Anthropic
Recommendation coverage
Where this model appears
- #1Best LLM for Agentic Coding
- #1Best LLM for Tool & Function Calling
- #1Best LLM for AI Agents
- #1Best LLM for Autonomous Agents
- #1Best LLM for Computer & Browser Use
- #1Best LLM for Hermes Agent
- #1Best LLM for OpenClaw
- #1Best LLM for Goose
- #1Best LLM for Coding
- #1Best LLM for OpenHands
- #2Best LLM for Frontend & UI
Compare head to head
Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.