BlogLLM Benchmark Scores as Published, Side by Side
Claude Fable 5: Benchmark Scores as Published
Every result the catalog holds for Claude Fable 5: 129 tests in 18 areas, from 21 pages. 36 printed by Anthropic itself, the rest by the company named beside each number. Each number is as printed; nothing is ranked or rescored.
Developer AnthropicReleased 2026-06-09Context window 1MMax output 128KInput Text, imageOutput TextWeights ClosedData version 2026103001, as of 2026-10-30
Knowledge Work
Agentic Coding
ML Engineering
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| PostTrainBench v1.1 | 41.4% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Harness: Claude Code; Fallback: w/ fallback |
| PaperBench | 88.8 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14 | - |
| MLS-Bench-Lite | 49.9 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14; printed on 1 more page | - |
| Internal Data Science Tasks | 34.7% | OpenAI, GPT-6 Astrachecked 2026-10-14 | - |
Science and Math
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPQA Diamond | 92.6% | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14; printed on 4 more pages | - |
| SciCode | 60.2% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
| CritPt | 28.6% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
| BioMysteryBench Human Difficult | 46.5% | Anthropic, Claude Opus 5checked 2026-10-15 | - |
| Terminal-Bench Science 0.1 | 24.7% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 5 more pages | - |
| FrontierMath Tier 4 (v2) | 87.8% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 1 more page | - |
| FrontierMath Tier 1-3 (v2) | 87% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
Long Context
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| AA-LCR | 70% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
Computer Use
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| OSWorld-Verified | 85% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Fallback: w/ fallback |
| OSWorld 2.0 | 66.1% | Anthropic, Claude Opus 5checked 2026-10-15; printed on 1 more page | - |
| OSWorld 2.0 partial | 72.9% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 1 more page | - |
| ScreenSpot-Pro | 87.3% | OpenAI, GPT-6 Astrachecked 2026-10-14 | Tools: none |
| OSWorld 2.0 strict | 36.1% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 1 more page | - |
Multimodal Understanding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| CharXiv Reasoning (no tools) | 88.9% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Tools: none; Fallback: w/ fallback |
| MMMU-Pro | 81.2% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Tools: none; Fallback: w/ fallback |
| MathVision | 94.8% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Tools: none; Fallback: w/ fallback |
| GDP.pdf | 29.8% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| CharXiv Reasoning (with tools) | 93.5% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Tools: enabled; Python; Fallback: w/ fallback |
| BabyVision (with tools) | 90.5% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Tools: enabled; Python; Fallback: w/ fallback |
| MMMU-Pro (with tools) | 86.5% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Tools: enabled; Python; Fallback: w/ fallback |
| Chartography (no tools) | 36.6% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | Effort: max; Tools: none |
| Chartography (with tools) | 84.2% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | Effort: max; Tools: enabled |
| MathVision (with tools) | 98.6% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Tools: enabled; Python; Fallback: w/ fallback |
| ZeroBench (with tools, pass@5) | 46% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Tools: enabled; Python; Fallback: w/ fallback |
| PerceptionBench | 57.2% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
| ZeroBench (pass@5) | 23% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Tools: none; Fallback: w/ fallback |
| OmniDocBench (version not stated) | 89.8% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
| WorldVQA ForceAnswer | 56.7% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
Cybersecurity
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| CyberGym | 83.1% | DeepSeek, DeepSeek-V4-Pro-0813 model cardchecked 2026-10-15; printed on 1 more page | - |
| ExploitBench | 78% | Xiaomi, Xiaomi MiMo-V2.6-Pro-RL model cardchecked 2026-10-19; printed on 1 more page | - |
| ExploitGym | 28.4% | OpenAI, GPT-6 Astrachecked 2026-10-14; printed on 1 more page | - |
| ExploitGym (2h) | 181 | Z.ai, GLM-5.3 model cardchecked 2026-10-15 | Attempt budget: timeout budget 2 hours (rescaled API inference time plus overhead); Fallback: w/ fallback |
| ExploitGym (6h) | 247 | Z.ai, GLM-5.3 model cardchecked 2026-10-15 | Attempt budget: timeout budget 6 hours (rescaled API inference time plus overhead); Fallback: w/ fallback |
General Knowledge and Reasoning
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Humanity's Last Exam (no tools) | 57.8% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 6 more pages | Tools: none |
| Humanity's Last Exam (with tools) | 63.8% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 8 more pages | Tools: enabled; exact tools in source |
| SimpleQA Verified | 68.3% | Thinking Machines, Thinking Machines Inkling model cardchecked 2026-10-15 | Effort: max |
| ARC-AGI-2 | 89.2% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25; printed on 1 more page | Effort: max |
| AA-Omniscience Index | 40 | Thinking Machines, Thinking Machines Inkling model cardchecked 2026-10-15 | Effort: max |
| ARC-AGI-1 | 98.5% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25; printed on 1 more page | Effort: max |
Code Generation
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| QwenReactBench | 1770 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14 | - |
| QwenSVGBench | 1690 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14 | - |
Tool Use and Workflows
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Toolathlon | 61.7% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| τ³-bench Banking | 26.8% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Fallback: w/ fallback |
| Agents' Last Exam (pass rate) | 25.7% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 2 more pages | Effort: xhigh; Harness: Claude Code; Fallback: w/ fallback; 40% of tasks annotated as downgraded (dagger on the card) |
| Toolathlon Verified | 77.9% | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14; printed on 4 more pages | - |
| MCP Atlas | 84.7% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: max; Fallback: w/ fallback |
| Agents' Last Exam (score) | 40.5% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 1 more page | - |
| SkillsBench | 70.9 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14 | - |
| MCPMark-Verified | 87.4% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
Web Search and Research
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| BrowseComp | 87.4% | Anthropic, Claude Opus 5checked 2026-10-15; printed on 3 more pages | - |
| WideSearch (metric not stated) | 81.2 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14 | - |
| DeepSearchQA (F1) | 94.2% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: max; Fallback: w/ fallback |
Instruction Following
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| IFBench | 63.5 | Alibaba, Qwen3.8-Max comparisonchecked 2026-10-14; printed on 1 more page | - |
Multilingual
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Global-MMLU | 93.6% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | Effort: max |
| Global-MMLU-Lite | 93.3% | Thinking Machines, Thinking Machines Inkling model cardchecked 2026-10-15 | Effort: max |
| MILU | 92.9% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | Effort: max |
Healthcare
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| HealthBench Professional (length-adjusted) | 63.3% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25; printed on 1 more page | Effort: max |
| HealthBench Professional | 60.9% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| HealthBench (unadjusted) | 61.2 | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | - |
| HealthBench Professional (unadjusted) | 68.9% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | - |
Safety and Alignment
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| FORTRESS (Adversarial) | 96% | Thinking Machines, Thinking Machines Inkling model cardchecked 2026-10-15 | Effort: max |
| FORTRESS (Benign) | 55.1% | Thinking Machines, Thinking Machines Inkling model cardchecked 2026-10-15 | Effort: max |
| StrongREJECT | 98.7% | Thinking Machines, Thinking Machines Inkling model cardchecked 2026-10-15 | Effort: max |
| Internal computer use safety benchmarklower is better | 18.3% | OpenAI, GPT-6 Astrachecked 2026-10-14 | - |
Engineering and Design
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| BenchCAD | 67.5% | OpenAI, GPT-6 Astrachecked 2026-10-14 | - |
| EEBench reward | 54.2% | xAI, Grok 4.7 model cardchecked 2026-10-14 | Effort: max |
| BenchCAD Vision2Code, no tools (voxel IoU) | 0.376 | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | Tools: none |
| BenchCAD Vision2Code, with tools (voxel IoU) | 0.675 | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25 | Tools: enabled |
| Internal Design Tasks | 35.8% | OpenAI, GPT-6 Astrachecked 2026-10-14 | - |
Preference and Composite Rankings
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Artificial Analysis Intelligence Index | 62 | xAI, Introducing Grok 4.6checked 2026-10-15 | Effort: max |
| Artificial Analysis Intelligence Index v4.1 | 59.9 | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| Artificial Analysis Intelligence Index v4.1.1 | 62.1 | OpenAI, GPT-6 Astrachecked 2026-10-14 | - |
Price
API list prices in US dollars per one million tokens, as printed on the developer's pricing page on the day checked. Cost per task is measured by Vals AI.
| Price | Value | Printed by | Setup as printed |
|---|---|---|---|
| Input | $10 | Anthropic, Anthropic API pricingchecked 2026-10-15 | - |
| Cached input | $1 | Anthropic, Anthropic API pricingchecked 2026-10-15 | - |
| Cache write | $12.5 | Anthropic, Anthropic API pricingchecked 2026-10-15 | Cache: 5 minute cache |
| Cache write, 1 hour | $20 | Anthropic, Anthropic API pricingchecked 2026-10-15 | - |
| Output | $50 | Anthropic, Anthropic API pricingchecked 2026-10-15 | - |
| Input, batch | $5 | Anthropic, Anthropic API pricingchecked 2026-10-15 | - |
| Output, batch | $25 | Anthropic, Anthropic API pricingchecked 2026-10-15 | - |
| Cost per taskVals AI, USD, lower is better | $41.71 | Vals AI, Vibe Code Bench v1.1 leaderboardchecked 2026-10-27 | Harness: OpenHands |
More from Anthropic
- Claude Haiku 5.5
- Claude Sonnet 5.5
- Claude Opus 5.5
- Claude Fable 5.1
- Claude Mythos 5.1
- Claude Opus 5
- Claude Sonnet 5
- Claude Mythos 5
- Claude Opus 4.8
- Claude Opus 4.7
- Claude Mythos Preview
- Claude Sonnet 4.6
- Claude Opus 4.6
- Claude Opus 4.5
- Claude Haiku 4.5
- Claude Sonnet 4.5
- Claude Opus 4.1
- Claude Opus 4
- Claude Sonnet 4
- Claude 3.7 Sonnet
- Claude 3.5 Sonnet
Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.