Model comparison
Claude Opus 4 vs Claude Sonnet 5
Claude Sonnet 5 is the stronger model overall, scoring 54.6 to 43.1 on the Noometry Index.
Last verified . 30 shared benchmarks.
Summary
- They share 30 benchmarks with published results for both. Claude Opus 4 scores higher in 1 category and Claude Sonnet 5 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Sonnet 5 leads 66.2 to 42.0.
- The biggest single-benchmark swing is GSO: 6.9% for Claude Opus 4 and 37.3% for Claude Sonnet 5.
- Claude Sonnet 5 is cheaper at $2 / $10 per million input/output tokens, against $15 / $75 for Claude Opus 4.
- Claude Sonnet 5 accepts more context: 1M tokens versus 200K.
Side by side
| Claude Opus 4 | Claude Sonnet 5 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 43.1 | 54.6 |
| Released | 2025-05-22 | 2026-06-29 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1M |
| Max output | 32K | 128K |
| Input $ / M tokens | $15 | $2 |
| Output $ / M tokens | $75 | $10 |
| Results tracked | 56 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 5 leads
Claude Opus 4: 47.2 (#62), Claude Sonnet 5: 55.5 (#26)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| GSO | 6.9% | 37.3% |
| WeirdML | 43.7% | 68.8% |
| LMArena Coding | 1442 | 1483 |
| SWE-bench Verified | 70.7% | — |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.7% |
| SWE-bench Verified (bash only) | 67.6% | — |
| Aider Polyglot | 72% | — |
| CursorBench | — | 34.1% |
| LMArena WebDev | — | 1541 |
| SciCode | — | 54.3% |
| ALE-Bench | — | 1,463 |
| AlgoTune | 1.33 | — |
Agentic & Tool Use Claude Sonnet 5 leads
Claude Opus 4: 34.8 (#42), Claude Sonnet 5: 42.8 (#18)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| LMArena Search | 1127 | 1194 |
| APEX-Agents | — | 54.5% |
| Cybench | 38% | — |
| DeepResearch Bench | 46.8% | — |
| GBAEval | — | 65.3% |
| METR Time Horizons | 63.9% | — |
| Vending-Bench 2 | — | 6,378 |
Reasoning Claude Sonnet 5 leads
Claude Opus 4: 27.3 (#121), Claude Sonnet 5: 49.1 (#39)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| SimpleBench | 58.8% | 60.6% |
| CritPt | 0.3% | 16.9% |
| LMArena Hard Prompts | 1399 | 1461 |
| DTBench | 81.6% | 92.5% |
| LMCA | 37.4% | 50% |
| Epoch Capabilities Index | 142.67 | 156.21 |
| ForecastBench | 61.1 | 61.1 |
| ARC-AGI-2 | 8.6% | — |
| Kagi LLM Benchmark | 74.3% | — |
| NYT Connections (extended) | — | 75.1% |
| ARC-AGI-1 | 35.7% | — |
| Chess Puzzles | — | 35% |
| EnigmaEval | 5.6% | — |
| Mystery Game Puzzles | — | 35% |
| Surface Evolver Bench | — | 60% |
| Bench to the Future 3 | — | 0.14 |
Math Claude Sonnet 5 leads
Claude Opus 4: 42.0 (#86), Claude Sonnet 5: 66.2 (#27)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 94.7% |
| LMArena Math | 1390 | 1467 |
| FrontierMath (Tiers 1-3) | — | 65.6% |
| FrontierMath Tier 4 | — | 29.3% |
| ProofBench | — | 77% |
| Omni-MATH | 61.6% | — |
| MATH Level 5 | 85% | — |
| FrontierMath (Feb 2025 set) | 4.5% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Claude Sonnet 5 leads
Claude Opus 4: 44.0 (#88), Claude Sonnet 5: 55.6 (#47)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| GPQA Diamond | 76.3% | 90.5% |
| LMArena Expert | 1386 | 1490 |
| Humanity's Last Exam | 10.7% | — |
| SimpleQA Verified | — | 33.7% |
| MMLU-Pro | 87.5% | — |
| Confabulations | 15.9% | — |
| Vectara Hallucination Rate | 12% | — |
| GPQA (HELM) | 70.8% | — |
Multimodal Claude Sonnet 5 leads
Claude Opus 4: 31.5 (#106), Claude Sonnet 5: 42.4 (#31)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| LMArena Vision | 1192 | 1274 |
| GeoBench | 49% | — |
| VPCT | 38% | — |
| Blueprint-Bench 2 | — | 24.9% |
| LMArena Document | — | 1466 |
Multilingual Claude Sonnet 5 leads
Claude Opus 4: 48.8 (#138), Claude Sonnet 5: 53.8 (#55)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| LMArena Non-English | 1362 | 1431 |
| LMArena Chinese | 1386 | 1477 |
| LMArena French | 1372 | 1460 |
| LMArena German | 1391 | 1440 |
| LMArena Japanese | 1331 | 1422 |
| LMArena Korean | 1321 | 1411 |
| LMArena Russian | 1392 | 1451 |
| LMArena Spanish | 1389 | 1437 |
Instruction Following Too close to call
Claude Opus 4: 77.1 (#28), Claude Sonnet 5: 76.3 (#41)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| LMArena Instruction Following | 1406 | 1452 |
| IFEval | 91.8% | — |
Long Context Claude Sonnet 5 leads
Claude Opus 4: 39.6 (#172), Claude Sonnet 5: 44.8 (#55)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| LMArena Longer Query | 1422 | 1463 |
| Fiction.LiveBench | 61.1% | — |
Writing & Preference Claude Sonnet 5 leads
Claude Opus 4: 61.2 (#89), Claude Sonnet 5: 69.2 (#25)
| Benchmark | Claude Opus 4 | Claude Sonnet 5 |
|---|---|---|
| LMArena Text | 1377 | 1442 |
| LMArena Creative Writing | 1387 | 1416 |
| EQ-Bench Creative Writing | 1580 | 1794 |
| LMArena Multi-Turn | 1396 | 1454 |
| Short-Story Creative Writing | 83.6% | — |
| WildBench | 85.2% | — |
| EQ-Bench 4 | — | 1236 |
Frequently asked questions
Is Claude Opus 4 better than Claude Sonnet 5?
Claude Sonnet 5 is the stronger model overall, scoring 54.6 to 43.1 on the Noometry Index.
Which is cheaper, Claude Opus 4 or Claude Sonnet 5?
Claude Sonnet 5 is cheaper. It lists at $2 per million input tokens and $10 per million output tokens; Claude Opus 4 lists at $15 and $75.
Is Claude Opus 4 or Claude Sonnet 5 better for coding?
Claude Sonnet 5 scores higher on coding benchmarks: 55.5 versus 47.2 in the Noometry coding category.
Which has the bigger context window?
Claude Sonnet 5 does, with 1M tokens against 200K.
How many benchmarks do Claude Opus 4 and Claude Sonnet 5 share?
30 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and Claude Sonnet 5 has 51.