# Claude Fable 5 vs Claude Sonnet 4

> Claude Fable 5 is the stronger model overall, scoring 66.3 to 40.6 on the Noometry Index. Claude Sonnet 4 costs 3.3× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/claude-fable-5-vs-claude-sonnet-4
- Last updated: 2026-10-10
- Shared benchmarks: 33

## Summary

- They share 33 benchmarks with published results for both. Claude Fable 5 scores higher in 10 categories and Claude Sonnet 4 in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Fable 5 leads 76.8 to 23.0.
- The biggest single-benchmark swing is ARC-AGI-2: 89.2% for Claude Fable 5 and 5.9% for Claude Sonnet 4.
- Claude Sonnet 4 is cheaper at $3 / $15 per million input/output tokens, against $10 / $50 for Claude Fable 5.
- Claude Fable 5 accepts more context: 1M tokens versus 200K.

## Snapshot

| | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 66.3 | 40.6 |
| Rank | 5 | 139 |
| Context | 1M | 200K |
| Input $/M | $10 | $3 |
| Output $/M | $50 | $15 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Fable 5: 67.0 (#4)
- Claude Sonnet 4: 44.0 (#70)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| SciCode | 61% | 40% |
| WeirdML | 91.9% | 46.1% |
| LMArena Coding | 1519 | 1414 |
| ALE-Bench | 2,041 | 655.35 |
| DeepSWE | 69.9% | — |
| FrontierCode | 53.5% | — |
| SWE-bench Verified (bash only) | — | 64.9% |
| Aider Polyglot | — | 61.3% |
| LMArena WebDev | 1626 | — |
| FrontierSWE | 47% | — |
| GSO | — | 4.9% |
| MirrorCode | 63.9% | — |

## Agentic & Tool Use

- Claude Fable 5: 54.1 (#2)
- Claude Sonnet 4: 38.3 (#29)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| APEX-Agents | 63.6% | — |
| Remote Labor Index | 16.1% | — |
| TheAgentCompany | — | 33.1% |
| τ²-bench Banking | 39.7% | — |
| Cybench | — | 35% |
| DeepResearch Bench | — | 46.6% |
| OSWorld | — | 43.9% |
| PostTrainBench | 41.8% | — |
| GBAEval | 74.5% | — |
| GDP.pdf | 30% | — |
| LMArena Search | 1230 | — |
| METR Time Horizons | — | 62% |
| Vending-Bench 2 | 5,680 | — |

## Reasoning

- Claude Fable 5: 76.8 (#4)
- Claude Sonnet 4: 23.0 (#169)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| ARC-AGI-2 | 89.2% | 5.9% |
| SimpleBench | 81.9% | 45.5% |
| Kagi LLM Benchmark | 91.4% | 73% |
| ARC-AGI-1 | 98.5% | 40% |
| CritPt | 28.6% | 0.3% |
| EnigmaEval | 39.3% | 3.1% |
| LMArena Hard Prompts | 1508 | 1372 |
| DTBench | 98.4% | 77.1% |
| LMCA | 61.1% | 29% |
| Epoch Capabilities Index | 163.6 | 141.69 |
| NYT Connections (extended) | 92.7% | — |
| Chess Puzzles | 41% | — |
| EBR-Bench | 39.5% | — |
| Mystery Game Puzzles | 52% | — |
| Surface Evolver Bench | 95% | — |
| Bench to the Future 3 | 0.13 | — |
| ForecastBench | — | 60.2 |

## Math

- Claude Fable 5: 88.4 (#5)
- Claude Sonnet 4: 41.5 (#71)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 71.1% |
| LMArena Math | 1519 | 1375 |
| FrontierMath (Tiers 1-3) | 87% | — |
| FrontierMath Tier 4 | 90.2% | — |
| ProofBench | 95% | — |
| Omni-MATH | — | 60.2% |
| MATH Level 5 | — | 84.4% |
| FrontierMath (Feb 2025 set) | — | 4.1% |
| FrontierMath Erdős | 0% | — |
| FrontierMath Tier 4 (v1) | — | 0% |

## Knowledge

- Claude Fable 5: 62.3 (#22)
- Claude Sonnet 4: 41.6 (#99)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| GPQA Diamond | 85.9% | 79.2% |
| LMArena Expert | 1534 | 1371 |
| Humanity's Last Exam | — | 7.8% |
| SimpleQA Verified | 70.7% | — |
| MMLU-Pro | — | 84.3% |
| Confabulations | — | 13.2% |
| Vectara Hallucination Rate | — | 10.3% |
| GPQA (HELM) | — | 70.6% |

## Multimodal

- Claude Fable 5: 45.6 (#14)
- Claude Sonnet 4: 25.6 (#116)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Vision | 1326 | 1191 |
| GeoBench | — | 37% |
| VPCT | — | 34% |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 35.8% | — |
| LMArena Document | 1496 | — |
| MindCube | — | 44.8% |

## Multilingual

- Claude Fable 5: 57.6 (#8)
- Claude Sonnet 4: 46.9 (#151)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Non-English | 1483 | 1333 |
| LMArena Chinese | 1546 | 1351 |
| LMArena French | 1509 | 1364 |
| LMArena German | 1486 | 1331 |
| LMArena Japanese | 1507 | 1301 |
| LMArena Korean | 1493 | 1293 |
| LMArena Russian | 1507 | 1355 |
| LMArena Spanish | 1503 | 1356 |

## Instruction Following

- Claude Fable 5: 78.8 (#7)
- Claude Sonnet 4: 71.8 (#141)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Instruction Following | 1503 | 1375 |
| IFEval | — | 84% |

## Long Context

- Claude Fable 5: 46.3 (#22)
- Claude Sonnet 4: 33.6 (#254)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Longer Query | 1510 | 1398 |
| Fiction.LiveBench | — | 46.9% |

## Writing & Preference

- Claude Fable 5: 76.1 (#5)
- Claude Sonnet 4: 57.3 (#127)

| Benchmark | Claude Fable 5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Text | 1491 | 1351 |
| LMArena Creative Writing | 1494 | 1345 |
| EQ-Bench Creative Writing | 1943 | 1483 |
| LMArena Multi-Turn | 1505 | 1376 |
| Short-Story Creative Writing | — | 81.4% |
| WildBench | — | 83.8% |
| EQ-Bench 4 | 1340 | — |

## FAQ

### Is Claude Fable 5 better than Claude Sonnet 4?

Claude Fable 5 is the stronger model overall, scoring 66.3 to 40.6 on the Noometry Index. Claude Sonnet 4 costs 3.3× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.

### Which is cheaper, Claude Fable 5 or Claude Sonnet 4?

Claude Sonnet 4 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; Claude Fable 5 lists at $10 and $50.

### Is Claude Fable 5 or Claude Sonnet 4 better for coding?

Claude Fable 5 scores higher on coding benchmarks: 67.0 versus 44.0 in the Noometry coding category.

### Which has the bigger context window?

Claude Fable 5 does, with 1M tokens against 200K.

### How many benchmarks do Claude Fable 5 and Claude Sonnet 4 share?

33 benchmarks have published results for both models. Claude Fable 5 has 61 scored results on Noometry and Claude Sonnet 4 has 58.