Claude Fable 5.1
| Released | 2026-09-01 |
|---|---|
| Price per 1M | $10.00 in · $50.00 out |
| Overall score | 65.3 · rank 5 |
| Context window | 1,000,000 tokens |
| Provider | Anthropic |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 65.3 | 5 | 55 |
| Overall: Coding | 65.2 | 5 | 40 |
| Overall: Reasoning | 65.8 | 7 | 19 |
| Overall: Agentic work | 64.5 | 6 | 37 |
| Overall: Frontend | 64.9 | 4 | 17 |
| Coding | 65.7 | 5 | 12 |
| Agents | 63.6 | 7 | 14 |
| Reasoning | 66.1 | 7 | 11 |
| Math | 66.4 | 6 | 4 |
| Knowledge | 64.2 | 8 | 4 |
| Frontend | 64.7 | 4 | 3 |
| General | 68.0 | 1 | 5 |
| Multimodal | 63.8 | 1 | 2 |
| Value vs price | +8.6 | 42 |
Buzz
| Posts on X | About 157 a day |
|---|---|
| Mood | 16% positive, 32% negative |
| Hacker News | 51 mentions this week |
| Checked | 2026-11-11 |
- Unverified rumours 😃 Physical AI anxiety and Huawei Ascend pretraining: Sources close to DeepSeek note that OpenAI's GPT 6 Astra hitting 95% accuracy on a robotic control task (versus 40% for Claude Fable 5.1) triggered internal concern that scaling text and code LLMs alone wi…
- Here’s proof, btw. This isn’t even frontier work yet, just some unusual thinking. I’ll test some frontier stuff throughout the rest of the week. I’m willing to go back to Fable 5.1 if needed; it’s still a good colleague. https://t.co/5n29wfZSDb
- Fable 5.2 declared “task complete in 30 trials. It was right once. Every false claim came after real progress, so the work looked finished, which makes it harder to catch. Fable 5.2 made more false claims than any other Claude model: 29, against 11 for Fable 5.1, 3 for Claude …
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 68.7 ±2.3 | 14 |
| High | 64.0 ±2.2 | 12 |
| Low | 62.2 ±2.7 | 6 |
| Max | 64.8 ±0.8 | 34 |
| Medium | 63.0 ±2.9 | 7 |
| Extra High | 64.1 ±2.2 | 8 |
Evidence
Coding 65.7 ±2.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| CursorBench | 51.8% | Epoch AI | 2026-11-11 |
| FrontierSWE | 56.3% | Epoch AI | 2026-11-11 |
| FrontierCode | 50.9% | Epoch AI | 2026-11-11 |
| LiveBench 2.0 Coding | 89.7% | LiveBench | 2026-11-06 |
| WeirdML v3 | 26.0% | Epoch AI | 2026-11-11 |
| MirrorCode | 93.3% | Epoch AI | 2026-09-09 |
| GSO | 88.2% | Epoch AI | 2026-11-11 |
| IOI | 90.8% | Vals AI | 2026-10-29 |
| SWE-bench Multimodal | 61.3% | Vals AI | 2026-11-09 |
| SciCode | 63.1% | Epoch AI | 2026-11-11 |
| LiveCodeBench (Vals run) | 90.5% | Vals AI | 2026-09-01 |
| ProgramBench | 7.0% | Vals AI | 2026-11-09 |
Agents 63.6 ±0.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 4 | 57.6% | Terminal-Bench | 2026-09-01 |
| SRE Bench | 22.9% | Vals AI | 2026-11-05 |
| Terminal-Bench 4.0 (Vals run) | 58.1% | Vals AI | 2026-11-09 |
| Terminal-Bench-Science | 40.0% | Vals AI | 2026-11-09 |
| APEX-Agents | 68.6% | Epoch AI | 2026-11-11 |
| Remote Labor Index | 17.9% | Epoch AI | 2026-11-11 |
| LiveBench 2.0 Agentic | 74.2% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Browser Use | 89.8% | LiveBench | 2026-11-06 |
| Terminal-Bench 2.1 (Vals run) | 85.0% | Vals AI | 2026-10-01 |
| EMB v1.1 (Excel financial modeling) | 76.7% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Research | 88.5% | LiveBench | 2026-11-06 |
| Tax Agent Bench | 50.7% | Vals AI | 2026-11-09 |
| SkillsBench | 61.6% | Vals AI | 2026-10-29 |
| Vending-Bench 2 | $5,422 | Epoch AI | 2026-11-11 |
| CyberBench v1.1 | 70.4% | Vals AI | 2026-11-05 |
Reasoning 66.1 ±1.4
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 90.0% | Epoch AI | 2026-11-11 |
| ARC-AGI-2 (ARC Prize verified) | 90.0% | ARC Prize | 2026-11-11 |
| Extended NYT Connections | 90.0% | Lech Mazur | 2026-11-10 |
| MysteryMechanism | 47.7% | Vals AI | 2026-11-09 |
| GPQA Diamond (Vals run) | 93.4% | Vals AI | 2026-09-01 |
| CritPt | 31.1% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 97.5% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 97.5% | Epoch AI | 2026-11-11 |
| SimpleBench | 86.6% | SimpleBench | 2026-09-03 |
| Chess Puzzles | 48.0% | Epoch AI | 2026-10-13 |
| LiveBench 2.0 Reasoning | 91.7% | LiveBench | 2026-11-06 |
| Mystery Game Puzzles | 58.0% | Epoch AI | 2026-09-01 |
| LiveBench 2.0 Data Analysis | 80.3% | LiveBench | 2026-11-06 |
Math 66.4 ±1.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 | 90.2% | Epoch AI | 2026-09-01 |
| FrontierMath Tier 4 | 87.8% | Epoch AI | 2026-09-01 |
| OTIS Mock AIME 2024-2025 | 100.0% | Epoch AI | 2026-09-01 |
| ProofBench v1.1 | 100.0% | Vals AI | 2026-11-09 |
| ProofBench | 100.0% | Epoch AI | 2026-11-11 |
Knowledge 64.2 ±2.4
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 70.8% | Epoch AI | 2026-09-01 |
| Humanity's Last Exam | 46.5% | Epoch AI | 2026-11-11 |
| MMLU-Pro (Vals run) | 92.4% | Vals AI | 2026-09-01 |
| Legal Research Bench | 55.3% | Vals AI | 2026-11-09 |
Frontend 64.7 ±1.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Design | 81.6% | LiveBench | 2026-11-06 |
| Vibe Code Bench v1.1 | 90.3% | Vals AI | 2026-11-09 |
| LMArena WebDev | 1,745 | LMArena | 2026-11-09 |
General 68.0 ±2.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LLM Debate Benchmark | 1,727 | Lech Mazur | 2026-10-26 |
| EQ-Bench Creative Writing v3 | 2,155 | EQ-Bench | 2026-11-11 |
| LMArena Text | 1,503 | LMArena | 2026-11-09 |
| Creative Story-Writing (Lech Mazur) | 93.0% | Lech Mazur | 2026-10-23 |
| EQ-Bench Longform Creative Writing | 85.3% | EQ-Bench | 2026-11-11 |
Multimodal 63.8 ±2.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMMU-Pro (Vals run) | 90.6% | Vals AI | 2026-09-01 |
| LMArena Vision | 1,289 | LMArena | 2026-11-09 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| APEX-Agents | 59.7% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 96.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 96.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 94.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 90.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 96.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 96.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 94.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 90.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 | 88.8% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 90.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 86.3% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 78.3% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 88.8% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 90.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 86.3% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 78.3% | ARC Prize | A stronger configuration result was used |
| CritPt | 29.7% | Epoch AI | A stronger configuration result was used |
| CritPt | 30.3% | Epoch AI | A stronger configuration result was used |
| CritPt | 29.1% | Epoch AI | A stronger configuration result was used |
| CritPt | 27.7% | Epoch AI | A stronger configuration result was used |
| CursorBench | 49.2% | Epoch AI | A stronger configuration result was used |
| CursorBench | 51.6% | Epoch AI | A stronger configuration result was used |
| CursorBench | 46.8% | Epoch AI | A stronger configuration result was used |
| CursorBench | 45.1% | Epoch AI | A stronger configuration result was used |
| MirrorCode | 73.3% | Epoch AI | A stronger configuration result was used |
| SciCode | 58.7% | Epoch AI | A stronger configuration result was used |
| SciCode | 60.9% | Epoch AI | A stronger configuration result was used |
| SciCode | 56.4% | Epoch AI | A stronger configuration result was used |
| SciCode | 56.7% | Epoch AI | A stronger configuration result was used |
| Terminal-Bench 4 | 57.3% | Terminal-Bench | A stronger configuration result was used |
| Terminal-Bench 4 | 54.9% | Terminal-Bench | A stronger configuration result was used |
| Terminal-Bench 4 | 53.9% | Terminal-Bench | A stronger configuration result was used |
| Terminal-Bench 4 | 43.3% | Terminal-Bench | A stronger configuration result was used |
Names matched to this model
claude-fable-5-1