AnthropicClaude Fable 5
Best configuration ranks #2 of 610 on the BenchLeader Index at 70.4 ±3.2. Last measured 8 Sept 2026. Released 9 Jun 2026.
- Blended price
- $20.00/M
- $10.00 in · $50.00 out
- Output speed
- 63 tok/s
- First answer
- 92 s
- first token 5.42 s
- Context
- 1M
- Overall index70
- Reasoning76
- Coding71
- Agents & tools74
- Knowledge71
- Instruction following63
- Human preference71
- Multimodal70
- Long context68
- Composite92
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| low | – | – | 63 tok/s | 92 s | $0.019 | Maths 67 · Reasoning 63 |
| medium | – | – | 63 tok/s | 92 s | $0.019 | Reasoning 69 |
| high | 63.9 | #40 | 63 tok/s | 92 s | $0.019 | Agents & tools 68 · Coding 76 · Maths 68 · Reasoning 67 |
| xhigh | – | – | 63 tok/s | 92 s | $0.019 | Knowledge 76 · Reasoning 72 |
| max | 65.8 | #23 | 63 tok/s | 92 s | $0.019 | Agents & tools 51 · Coding 75 · Composite 76 · Maths 76 · Reasoning 71 |
| thinking | – | – | 63 tok/s | 92 s | $0.019 | Reasoning 82 |
| defaultbest | 70.4 | #2 | 63 tok/s | 92 s | $0.019 | Agents & tools 74 · Coding 71 · Composite 92 · Human preference 71 · Instruction following 63 · Knowledge 71 · Long context 68 · Multimodal 70 · Reasoning 76 |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| GPQA Diamond | 78.8%#120 | – | 83.3%#91 | – | 85.9%#72 | – | – | Epoch AI Benchmarking Hub | |
| SimpleBench | – | – | – | – | 81.9%#1 | – | – | SimpleBench | |
| LMArena Hard Prompts | – | – | – | – | – | – | 1533#1 | LMArena | |
| LiveBench Reasoningnot in index | – | – | – | – | 89.7%#9 | – | – | LiveBench | |
| GPQA Diamond (AA)not in index | – | – | – | – | – | – | 92.6%#32 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | – | – | – | – | – | 55.5%#4 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | – | – | – | – | – | – | 93.2%#10 | Vals AI | |
| EnigmaEval | – | – | 39.3%#1 | – | – | – | – | Scale AI SEAL | |
| Kagi LLM Benchmark | – | – | – | – | – | 91.4%#1 | 88.8%#2 | Kagi LLM Benchmark | |
| ARC-AGI-1 | 90.5%#50 | 92.5%#35 | 95.5%#20 | 98.5%#1 | 98.5%#1 | – | – | ARC Prize | |
| ARC-AGI-2 | 76.8%#29 | 82.5%#26 | 87.5%#14 | 88.3%#12 | 89.2%#10 | – | – | ARC Prize |
Coding
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| SciCode | – | – | – | – | 60.2%#2 | – | – | SciCode | |
| WeirdML | – | – | 87.8%#9 | – | 91.9%#4 | – | – | WeirdML | |
| FrontierCode | – | – | – | – | – | – | 53.5%#1 | Cognition | |
| LMArena Coding | – | – | – | – | – | – | 1551#3 | LMArena | |
| LMArena WebDev | – | – | – | – | – | – | 1628#9 | LMArena | |
| LiveBench Codingnot in index | – | – | – | – | 86.0%#2 | – | – | LiveBench | |
| SciCode (AA)not in index | – | – | – | – | – | – | 61.0%#2 | Artificial Analysis | |
| LiveCodeBench | – | – | – | – | – | – | 89.8%#2 | Vals AI | |
| IOI | – | – | – | – | – | – | 72.3%#5 | Vals AI | |
| SWE-bench (Vals)not in index | – | – | – | – | – | – | 95.0%#7 | Vals AI |
Agents & tools
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| Terminal-Bench | – | – | – | – | 44.5%#32 | – | – | Terminal-Bench | |
| Remote Labor Index | – | – | – | – | – | – | 16.1%#1 | Scale AI / CAIS | |
| APEX-Agents | – | – | – | – | – | – | 45.0%#3 | Mercor | |
| LMArena Agent | – | – | 9.2#5 | – | – | – | – | LMArena | |
| LiveBench Agentic Codingnot in index | – | – | – | – | 62.2%#6 | – | – | LiveBench | |
| Terminal-Bench Hard | – | – | – | – | – | – | 62.9%#2 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | – | – | – | – | – | – | 98.5%#4 | Artificial Analysis | |
| Terminal-Bench 2.1 (Vals) | – | – | – | – | – | – | 80.5%#7 | Vals AI | |
| MCP Atlas | – | – | – | – | – | – | 83.3%#5 | Scale AI SEAL | |
| HiL-Bench | – | – | – | – | – | – | 56.3%#3 | Scale AI SEAL |
Maths
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| FrontierMath Tiers 1–3 | – | – | – | – | 87.0%#5 | – | – | Epoch AI Benchmarking Hub | |
| FrontierMath Tier 4 | – | – | – | – | 90.2%#5 | – | – | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 97.8%#17 | – | 100.0%#1 | – | 99.7%#8 | – | – | Epoch AI Benchmarking Hub | |
| ProofBench | – | – | – | – | 95.0%#4 | – | – | Vals AI | |
| LiveBench Mathematicsnot in index | – | – | – | – | 96.0%#4 | – | – | LiveBench |
Knowledge
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| SimpleQA Verified | – | – | – | 70.7%#4 | – | – | – | Epoch AI Benchmarking Hub | |
| LiveBench Data Analysisnot in index | – | – | – | – | 80.5%#3 | – | – | LiveBench | |
| AA-Omniscience | – | – | – | – | – | – | 43.3#5 | Artificial Analysis | |
| MMLU-Pro | – | – | – | – | – | – | 91.5%#3 | Vals AI | |
| LegalBench | – | – | – | – | – | – | 88.6%#1 | Vals AI | |
| CorpFin | – | – | – | – | – | – | 71.8%#2 | Vals AI | |
| TaxEval | – | – | – | – | – | – | 76.9%#5 | Vals AI | |
| PRBench Finance | – | – | – | – | – | – | 53.9%#2 | Scale AI SEAL | |
| PRBench Legal | – | – | – | – | – | – | 52.6%#2 | Scale AI SEAL |
Instruction following
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| LiveBench Languagenot in index | – | – | – | – | 90.7%#1 | – | – | LiveBench | |
| IFBench | – | – | – | – | – | – | 63.5%#112 | Artificial Analysis |
Human preference
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| LMArena Text | – | – | – | – | – | – | 1507#1 | LMArena | |
| EQ-Bench 4 | – | – | – | – | – | – | 1340#2 | EQ-Bench |
Multimodal
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| LMArena Vision | – | – | – | – | – | – | 1330#1 | LMArena |
Long context
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| AA-LCR | – | – | – | – | – | – | 82.3%#21 | Artificial Analysis |
Composite
| Benchmark | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | – | – | – | – | – | 163.4#3 | Epoch AI Benchmarking Hub | |
| LiveBench | – | – | – | – | 83.0%#2 | – | – | LiveBench | |
| AA Intelligence Index | – | – | – | – | – | – | 49.7#8 | Artificial Analysis | |
| Vals Indexnot in index | – | – | – | – | – | – | 66.0#4 | Vals AI |
Where it wins
Benchmarks where this configuration ranks in the top five of every configuration measured.
- #1Remote Labor Index16.1% · of 12
- #1FrontierCode53.5% · of 29
- #1LMArena Text1507 · of 350
- #1LMArena Hard Prompts1533 · of 350
- #1LMArena Vision1330 · of 141
- #1LegalBench88.6% · of 136
- #2Terminal-Bench Hard62.9% · of 387
- #2LiveCodeBench89.8% · of 135
- #2CorpFin71.8% · of 125
- #2PRBench Finance53.9% · of 33
- #2PRBench Legal52.6% · of 33
- #2EQ-Bench 41340 · of 27
- #2Kagi LLM Benchmark88.8% · of 135
- #3APEX-Agents45.0% · of 63
- #3LMArena Coding1551 · of 345
- #3MMLU-Pro91.5% · of 129
- #3HiL-Bench56.3% · of 17
- #5AA-Omniscience43.3 · of 472
- #5IOI72.3% · of 58
- #5TaxEval76.9% · of 137
- #5MCP Atlas83.3% · of 30
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Google Vertex | 64 tok/s | 3.86 s | $10.00 | $50.00 | 1M | – |
| Anthropic | 44 tok/s | 5.45 s | $10.00 | $50.00 | 1M | – |
| Azure (BYOK Only) | 35 tok/s | 8.56 s | $10.00 | $50.00 | 1M | – |
| Claude Platform on AWS | 22 tok/s | 5.40 s | $10.00 | $50.00 | 1M | – |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $1.00 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.019 | $0.016 | 1.6 min |
| Summarise a 30-page report | 12,000 / 600 | $0.150 | $0.069 | 1.7 min |
| Code edit | 6,000 / 1,500 | $0.135 | $0.095 | 1.9 min |
| Agentic coding session | 60,000 / 4,000 | $0.800 | $0.395 | 2.6 min |
| Structured extraction | 2,000 / 200 | $0.030 | $0.017 | 1.6 min |
See also
Data as of 9 Sept 2026. Compare these configurations.