Model comparison
Claude Opus 4 vs Claude Sonnet 4.6
Claude Sonnet 4.6 is the stronger model overall, scoring 50.3 to 43.1 on the Noometry Index.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. Claude Opus 4 scores higher in 1 category and Claude Sonnet 4.6 in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Sonnet 4.6 leads 46.1 to 27.3.
- The biggest single-benchmark swing is ARC-AGI-2: 8.6% for Claude Opus 4 and 60.4% for Claude Sonnet 4.6.
- Claude Sonnet 4.6 is cheaper at $3 / $15 per million input/output tokens, against $15 / $75 for Claude Opus 4.
- Claude Sonnet 4.6 accepts more context: 1M tokens versus 200K.
Side by side
| Claude Opus 4 | Claude Sonnet 4.6 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 43.1 | 50.3 |
| Released | 2025-05-22 | 2026-02-17 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1M |
| Max output | 32K | 128K |
| Input $ / M tokens | $15 | $3 |
| Output $ / M tokens | $75 | $15 |
| Results tracked | 56 | 57 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Claude Opus 4: 47.2 (#62), Claude Sonnet 4.6: 46.3 (#67)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| SWE-bench Verified | 70.7% | 75.2% |
| WeirdML | 43.7% | 66.1% |
| LMArena Coding | 1442 | 1504 |
| DeepSWE | — | 29.9% |
| FrontierCode | — | 24.3% |
| SWE-bench Verified (bash only) | 67.6% | — |
| Aider Polyglot | 72% | — |
| LMArena WebDev | — | 1522 |
| SciCode | — | 46.8% |
| GSO | 6.9% | — |
| ALE-Bench | — | 1,327 |
| AlgoTune | 1.33 | — |
Agentic & Tool Use Claude Sonnet 4.6 leads
Claude Opus 4: 34.8 (#42), Claude Sonnet 4.6: 39.1 (#28)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| DeepResearch Bench | 46.8% | 54.9% |
| LMArena Search | 1127 | 1221 |
| Terminal-Bench | — | 53.4% |
| APEX-Agents | — | 43% |
| OSWorld 2.0 | — | 9.3% |
| Cybench | 38% | — |
| OSWorld | — | 72.1% |
| ExploitBench | — | 23.6% |
| GBAEval | — | 48.8% |
| GDP.pdf | — | 18% |
| METR Time Horizons | 63.9% | — |
| Vending-Bench 2 | — | 7,204 |
Reasoning Claude Sonnet 4.6 leads
Claude Opus 4: 27.3 (#121), Claude Sonnet 4.6: 46.1 (#45)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| ARC-AGI-2 | 8.6% | 60.4% |
| ARC-AGI-1 | 35.7% | 86.5% |
| CritPt | 0.3% | 3.1% |
| LMArena Hard Prompts | 1399 | 1484 |
| DTBench | 81.6% | 89.9% |
| LMCA | 37.4% | 46.5% |
| Epoch Capabilities Index | 142.67 | 152.24 |
| ForecastBench | 61.1 | 62 |
| SimpleBench | 58.8% | — |
| Kagi LLM Benchmark | 74.3% | — |
| NYT Connections (extended) | — | 80.9% |
| Chess Puzzles | — | 13% |
| EnigmaEval | 5.6% | — |
| Thematic Generalization | — | 76.3% |
| Mystery Game Puzzles | — | 16% |
Math Claude Sonnet 4.6 leads
Claude Opus 4: 42.0 (#86), Claude Sonnet 4.6: 52.9 (#49)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 85.8% |
| LMArena Math | 1390 | 1462 |
| FrontierMath (Feb 2025 set) | 4.5% | 32.4% |
| FrontierMath Tier 4 (v1) | 4.2% | 8.3% |
| ProofBench | — | 45% |
| Omni-MATH | 61.6% | — |
| MATH Level 5 | 85% | — |
Knowledge Claude Sonnet 4.6 leads
Claude Opus 4: 44.0 (#88), Claude Sonnet 4.6: 51.7 (#65)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| GPQA Diamond | 76.3% | 87.4% |
| Vectara Hallucination Rate | 12% | 10.6% |
| LMArena Expert | 1386 | 1500 |
| Humanity's Last Exam | 10.7% | — |
| SimpleQA Verified | — | 35.5% |
| MMLU-Pro | 87.5% | — |
| Confabulations | 15.9% | — |
| GPQA (HELM) | 70.8% | — |
Multimodal Claude Sonnet 4.6 leads
Claude Opus 4: 31.5 (#106), Claude Sonnet 4.6: 38.0 (#68)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| LMArena Vision | 1192 | 1283 |
| GeoBench | 49% | — |
| VPCT | 38% | — |
| Blueprint-Bench 2 | — | 6.7% |
| LMArena Document | — | 1482 |
Multilingual Claude Sonnet 4.6 leads
Claude Opus 4: 48.8 (#138), Claude Sonnet 4.6: 54.4 (#41)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| LMArena Non-English | 1362 | 1440 |
| LMArena Chinese | 1386 | 1491 |
| LMArena French | 1372 | 1465 |
| LMArena German | 1391 | 1428 |
| LMArena Japanese | 1331 | 1420 |
| LMArena Korean | 1321 | 1411 |
| LMArena Russian | 1392 | 1440 |
| LMArena Spanish | 1389 | 1464 |
Instruction Following Too close to call
Claude Opus 4: 77.1 (#28), Claude Sonnet 4.6: 77.4 (#25)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| LMArena Instruction Following | 1406 | 1475 |
| IFEval | 91.8% | — |
Long Context Claude Sonnet 4.6 leads
Claude Opus 4: 39.6 (#172), Claude Sonnet 4.6: 45.3 (#44)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| LMArena Longer Query | 1422 | 1479 |
| Fiction.LiveBench | 61.1% | — |
Writing & Preference Claude Sonnet 4.6 leads
Claude Opus 4: 61.2 (#89), Claude Sonnet 4.6: 70.2 (#22)
| Benchmark | Claude Opus 4 | Claude Sonnet 4.6 |
|---|---|---|
| LMArena Text | 1377 | 1458 |
| LMArena Creative Writing | 1387 | 1435 |
| EQ-Bench Creative Writing | 1580 | 1810 |
| LMArena Multi-Turn | 1396 | 1464 |
| Short-Story Creative Writing | 83.6% | — |
| WildBench | 85.2% | — |
| EQ-Bench 4 | — | 1207 |
Frequently asked questions
Is Claude Opus 4 better than Claude Sonnet 4.6?
Claude Sonnet 4.6 is the stronger model overall, scoring 50.3 to 43.1 on the Noometry Index.
Which is cheaper, Claude Opus 4 or Claude Sonnet 4.6?
Claude Sonnet 4.6 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; Claude Opus 4 lists at $15 and $75.
Is Claude Opus 4 or Claude Sonnet 4.6 better for coding?
They score almost the same on coding (47.2 vs 46.3); test both on your own repository before choosing.
Which has the bigger context window?
Claude Sonnet 4.6 does, with 1M tokens against 200K.
How many benchmarks do Claude Opus 4 and Claude Sonnet 4.6 share?
35 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and Claude Sonnet 4.6 has 57.