Model comparison
Claude Opus 4 vs Claude Opus 5.5
Claude Opus 5.5 is the stronger model overall, scoring 68.6 to 43.1 on the Noometry Index.
Last verified . 24 shared benchmarks.
Summary
- They share 24 benchmarks with published results for both. Claude Opus 4 scores higher in 0 categories and Claude Opus 5.5 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 5.5 leads 80.2 to 27.3.
- The biggest single-benchmark swing is ARC-AGI-2: 8.6% for Claude Opus 4 and 93.3% for Claude Opus 5.5.
- Claude Opus 5.5 is cheaper at $4 / $20 per million input/output tokens, against $15 / $75 for Claude Opus 4.
- Claude Opus 5.5 accepts more context: 1M tokens versus 200K.
Side by side
| Claude Opus 4 | Claude Opus 5.5 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 43.1 | 68.6 |
| Released | 2025-05-22 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1M |
| Max output | 32K | 128K |
| Input $ / M tokens | $15 | $4 |
| Output $ / M tokens | $75 | $20 |
| Results tracked | 56 | 44 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 5.5 leads
Claude Opus 4: 47.2 (#62), Claude Opus 5.5: 71.9 (#3)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| LMArena Coding | 1442 | 1547 |
| SWE-bench Verified | 70.7% | — |
| FrontierCode | — | 54.6% |
| SWE-bench Verified (bash only) | 67.6% | — |
| Aider Polyglot | 72% | — |
| CursorBench | — | 57.8% |
| LMArena WebDev | — | 1813 |
| FrontierSWE | — | 62.3% |
| SciCode | — | 66.9% |
| GSO | 6.9% | — |
| WeirdML | 43.7% | — |
| MirrorCode | — | 77.4% |
| ALE-Bench | — | 2,147 |
| AlgoTune | 1.33 | — |
Agentic & Tool Use Claude Opus 5.5 leads
Claude Opus 4: 34.8 (#42), Claude Opus 5.5: 45.3 (#15)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| APEX-Agents | — | 73.5% |
| Cybench | 38% | — |
| DeepResearch Bench | 46.8% | — |
| GDP.pdf | — | 30.6% |
| LMArena Search | 1127 | — |
| METR Time Horizons | 63.9% | — |
| Vending-Bench 2 | — | 9,235 |
Reasoning Claude Opus 5.5 leads
Claude Opus 4: 27.3 (#121), Claude Opus 5.5: 80.2 (#3)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| ARC-AGI-2 | 8.6% | 93.3% |
| ARC-AGI-1 | 35.7% | 98.5% |
| CritPt | 0.3% | 31.7% |
| LMArena Hard Prompts | 1399 | 1535 |
| DTBench | 81.6% | 98.9% |
| LMCA | 37.4% | 68.2% |
| Epoch Capabilities Index | 142.67 | 167.33 |
| SimpleBench | 58.8% | — |
| Kagi LLM Benchmark | 74.3% | — |
| NYT Connections (extended) | — | 88.5% |
| EnigmaEval | 5.6% | — |
| EBR-Bench | — | 71.4% |
| Mystery Game Puzzles | — | 71% |
| ForecastBench | 61.1 | — |
Math Claude Opus 5.5 leads
Claude Opus 4: 42.0 (#86), Claude Opus 5.5: 91.8 (#3)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 100% |
| LMArena Math | 1390 | 1506 |
| FrontierMath (Tiers 1-3) | — | 91.2% |
| FrontierMath Tier 4 | — | 95% |
| ProofBench | — | 100% |
| Omni-MATH | 61.6% | — |
| MATH Level 5 | 85% | — |
| FrontierMath (Feb 2025 set) | 4.5% | — |
| FrontierMath Erdős | — | 2.9% |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Claude Opus 5.5 leads
Claude Opus 4: 44.0 (#88), Claude Opus 5.5: 66.4 (#10)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| GPQA Diamond | 76.3% | 90.6% |
| LMArena Expert | 1386 | 1547 |
| Humanity's Last Exam | 10.7% | — |
| SimpleQA Verified | — | 72.2% |
| MMLU-Pro | 87.5% | — |
| Confabulations | 15.9% | — |
| Vectara Hallucination Rate | 12% | — |
| GPQA (HELM) | 70.8% | — |
Multimodal Claude Opus 5.5 leads
Claude Opus 4: 31.5 (#106), Claude Opus 5.5: 57.8 (#1)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| LMArena Vision | 1192 | 1321 |
| GeoBench | 49% | — |
| VPCT | 38% | — |
| Blueprint-Bench 2 | — | 51.2% |
| Furniture Assembly | — | 83.3% |
Multilingual Claude Opus 5.5 leads
Claude Opus 4: 48.8 (#138), Claude Opus 5.5: 59.1 (#2)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| LMArena Non-English | 1362 | 1507 |
| LMArena Chinese | 1386 | 1588 |
| LMArena French | 1372 | 1514 |
| LMArena Russian | 1392 | 1520 |
| LMArena Spanish | 1389 | 1507 |
| LMArena German | 1391 | — |
| LMArena Japanese | 1331 | — |
| LMArena Korean | 1321 | — |
Instruction Following Claude Opus 5.5 leads
Claude Opus 4: 77.1 (#28), Claude Opus 5.5: 80.0 (#3)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| LMArena Instruction Following | 1406 | 1537 |
| IFEval | 91.8% | — |
Long Context Claude Opus 5.5 leads
Claude Opus 4: 39.6 (#172), Claude Opus 5.5: 47.1 (#19)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| LMArena Longer Query | 1422 | 1532 |
| Fiction.LiveBench | 61.1% | — |
Writing & Preference Claude Opus 5.5 leads
Claude Opus 4: 61.2 (#89), Claude Opus 5.5: 78.2 (#3)
| Benchmark | Claude Opus 4 | Claude Opus 5.5 |
|---|---|---|
| LMArena Text | 1377 | 1515 |
| LMArena Creative Writing | 1387 | 1533 |
| EQ-Bench Creative Writing | 1580 | 2050 |
| LMArena Multi-Turn | 1396 | 1499 |
| Short-Story Creative Writing | 83.6% | — |
| WildBench | 85.2% | — |
Frequently asked questions
Is Claude Opus 4 better than Claude Opus 5.5?
Claude Opus 5.5 is the stronger model overall, scoring 68.6 to 43.1 on the Noometry Index.
Which is cheaper, Claude Opus 4 or Claude Opus 5.5?
Claude Opus 5.5 is cheaper. It lists at $4 per million input tokens and $20 per million output tokens; Claude Opus 4 lists at $15 and $75.
Is Claude Opus 4 or Claude Opus 5.5 better for coding?
Claude Opus 5.5 scores higher on coding benchmarks: 71.9 versus 47.2 in the Noometry coding category.
Which has the bigger context window?
Claude Opus 5.5 does, with 1M tokens against 200K.
How many benchmarks do Claude Opus 4 and Claude Opus 5.5 share?
24 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and Claude Opus 5.5 has 44.