Model comparison
Claude Fable 5 vs Claude Sonnet 5
Claude Fable 5 is the stronger model overall, scoring 66.8 to 54.6 on the Noometry Index. Claude Sonnet 5 costs 5.0× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.
Last verified . 49 shared benchmarks.
Summary
- They share 49 benchmarks with published results for both. Claude Fable 5 scores higher in 10 categories and Claude Sonnet 5 in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Fable 5 leads 76.8 to 49.1.
- The biggest single-benchmark swing is FrontierMath Tier 4: 90.2% for Claude Fable 5 and 29.3% for Claude Sonnet 5.
- Claude Sonnet 5 is cheaper at $2 / $10 per million input/output tokens, against $10 / $50 for Claude Fable 5.
Side by side
| Claude Fable 5 | Claude Sonnet 5 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 66.8 | 54.6 |
| Released | 2026-06-07 | 2026-06-29 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1M |
| Max output | 128K | 128K |
| Input $ / M tokens | $10 | $2 |
| Output $ / M tokens | $50 | $10 |
| Results tracked | 62 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Fable 5 leads
Claude Fable 5: 70.6 (#4), Claude Sonnet 5: 55.5 (#26)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| DeepSWE | 69.9% | 53.8% |
| FrontierCode | 53.5% | 42.7% |
| LMArena WebDev | 1625 | 1541 |
| SciCode | 61% | 54.3% |
| GSO | 78.4% | 37.3% |
| WeirdML | 91.9% | 68.8% |
| LMArena Coding | 1519 | 1483 |
| ALE-Bench | 2,041 | 1,463 |
| CursorBench | — | 34.1% |
| FrontierSWE | 47% | — |
| MirrorCode | 63.9% | — |
Agentic & Tool Use Claude Fable 5 leads
Claude Fable 5: 54.0 (#2), Claude Sonnet 5: 42.8 (#18)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| APEX-Agents | 63.6% | 54.5% |
| GBAEval | 74.5% | 65.3% |
| LMArena Search | 1230 | 1194 |
| Vending-Bench 2 | 5,680 | 6,378 |
| Remote Labor Index | 16.1% | — |
| τ²-bench Banking | 39.7% | — |
| PostTrainBench | 41.8% | — |
| GDP.pdf | 30% | — |
Reasoning Claude Fable 5 leads
Claude Fable 5: 76.8 (#6), Claude Sonnet 5: 49.1 (#39)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| SimpleBench | 81.9% | 60.6% |
| NYT Connections (extended) | 92.7% | 75.1% |
| CritPt | 28.6% | 16.9% |
| Chess Puzzles | 41% | 35% |
| LMArena Hard Prompts | 1508 | 1461 |
| Mystery Game Puzzles | 52% | 35% |
| DTBench | 98.4% | 92.5% |
| LMCA | 61.1% | 50% |
| Surface Evolver Bench | 95% | 60% |
| Bench to the Future 3 | 0.13 | 0.14 |
| Epoch Capabilities Index | 162.06 | 156.21 |
| ARC-AGI-2 | 89.2% | — |
| Kagi LLM Benchmark | 91.4% | — |
| ARC-AGI-1 | 98.5% | — |
| EnigmaEval | 39.3% | — |
| EBR-Bench | 39.5% | — |
| ForecastBench | — | 61.1 |
Math Claude Fable 5 leads
Claude Fable 5: 88.5 (#5), Claude Sonnet 5: 66.2 (#27)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 87% | 65.6% |
| FrontierMath Tier 4 | 90.2% | 29.3% |
| OTIS Mock AIME 2024-2025 | 100% | 94.7% |
| ProofBench | 95% | 77% |
| LMArena Math | 1519 | 1467 |
| FrontierMath Erdős | 0% | — |
Knowledge Claude Fable 5 leads
Claude Fable 5: 62.2 (#25), Claude Sonnet 5: 55.6 (#47)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| GPQA Diamond | 85.9% | 90.5% |
| SimpleQA Verified | 70.7% | 33.7% |
| LMArena Expert | 1534 | 1490 |
Multimodal Claude Fable 5 leads
Claude Fable 5: 45.3 (#17), Claude Sonnet 5: 42.4 (#31)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| LMArena Vision | 1324 | 1274 |
| Blueprint-Bench 2 | 38.6% | 24.9% |
| LMArena Document | 1496 | 1466 |
| Furniture Assembly | 35.8% | — |
Multilingual Claude Fable 5 leads
Claude Fable 5: 57.3 (#9), Claude Sonnet 5: 53.8 (#55)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| LMArena Non-English | 1481 | 1431 |
| LMArena Chinese | 1543 | 1477 |
| LMArena French | 1505 | 1460 |
| LMArena German | 1486 | 1440 |
| LMArena Japanese | 1506 | 1422 |
| LMArena Korean | 1488 | 1411 |
| LMArena Russian | 1504 | 1451 |
| LMArena Spanish | 1498 | 1437 |
Instruction Following Claude Fable 5 leads
Claude Fable 5: 78.6 (#8), Claude Sonnet 5: 76.3 (#41)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| LMArena Instruction Following | 1502 | 1452 |
Long Context Claude Fable 5 leads
Claude Fable 5: 46.3 (#23), Claude Sonnet 5: 44.8 (#55)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| LMArena Longer Query | 1509 | 1463 |
Writing & Preference Claude Fable 5 leads
Claude Fable 5: 75.9 (#5), Claude Sonnet 5: 69.2 (#25)
| Benchmark | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| LMArena Text | 1491 | 1442 |
| LMArena Creative Writing | 1494 | 1416 |
| EQ-Bench Creative Writing | 1943 | 1794 |
| EQ-Bench 4 | 1340 | 1236 |
| LMArena Multi-Turn | 1504 | 1454 |
Frequently asked questions
Is Claude Fable 5 better than Claude Sonnet 5?
Claude Fable 5 is the stronger model overall, scoring 66.8 to 54.6 on the Noometry Index. Claude Sonnet 5 costs 5.0× 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 5?
Claude Sonnet 5 is cheaper. It lists at $2 per million input tokens and $10 per million output tokens; Claude Fable 5 lists at $10 and $50.
Is Claude Fable 5 or Claude Sonnet 5 better for coding?
Claude Fable 5 scores higher on coding benchmarks: 70.6 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do Claude Fable 5 and Claude Sonnet 5 share?
49 benchmarks have published results for both models. Claude Fable 5 has 62 scored results on Noometry and Claude Sonnet 5 has 51.