Model comparison
Claude Opus 5 vs GPT-5.6 Sol
Claude Opus 5 is the stronger model overall, scoring 67.8 to 65.0 on the Noometry Index.
Last verified . 57 shared benchmarks.
Summary
- They share 57 benchmarks with published results for both. Claude Opus 5 scores higher in 10 categories and GPT-5.6 Sol in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Claude Opus 5 leads 79.2 to 73.3.
- The biggest single-benchmark swing is GBAEval: 79.6% for Claude Opus 5 and 52.6% for GPT-5.6 Sol.
- GPT-5.6 Sol is cheaper at $4 / $20 per million input/output tokens, against $5 / $25 for Claude Opus 5.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Opus 5 | GPT-5.6 Sol | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 67.8 | 65.0 |
| Released | 2026-07-24 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $5 | $4 |
| Output $ / M tokens | $25 | $20 |
| Results tracked | 57 | 65 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 5 leads
Claude Opus 5: 67.5 (#5), GPT-5.6 Sol: 65.1 (#7)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| DeepSWE | 73.6% | 72.7% |
| FrontierCode | 53.4% | 47.5% |
| CursorBench | 46.6% | 41.7% |
| LMArena WebDev | 1691 | 1618 |
| FrontierSWE | 52% | 32.2% |
| SciCode | 56.4% | 57.1% |
| WeirdML | 91.8% | 89.4% |
| LMArena Coding | 1534 | 1498 |
| ALE-Bench | 2,165 | 2,177 |
| GSO | — | 76.5% |
| MirrorCode | — | 20% |
Agentic & Tool Use Claude Opus 5 leads
Claude Opus 5: 55.6 (#1), GPT-5.6 Sol: 50.3 (#7)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| APEX-Agents | 65.8% | 51.4% |
| OSWorld 2.0 | 31.4% | 27.3% |
| τ²-bench Banking | 48.7% | 46.9% |
| PostTrainBench | 35% | 36.2% |
| BALROG | 63.4% | 60% |
| GBAEval | 79.6% | 52.6% |
| GDP.pdf | 24% | 30.7% |
| Vending-Bench 2 | 11,182 | 9,619 |
| LMArena Search | — | 1257 |
Reasoning Claude Opus 5 leads
Claude Opus 5: 77.2 (#4), GPT-5.6 Sol: 74.8 (#8)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| ARC-AGI-2 | 90.4% | 92.5% |
| SimpleBench | 80.6% | 71.7% |
| NYT Connections (extended) | 94.3% | 93.8% |
| ARC-AGI-1 | 97.5% | 97.5% |
| CritPt | 29.1% | 32.3% |
| Chess Puzzles | 42% | 64% |
| EBR-Bench | 45.7% | 44.8% |
| LMArena Hard Prompts | 1526 | 1484 |
| Mystery Game Puzzles | 59% | 58% |
| DTBench | 97.9% | 96% |
| LMCA | 64.5% | 59.2% |
| Bench to the Future 3 | 0.12 | 0.14 |
| Epoch Capabilities Index | 162.78 | 161.66 |
| Kagi LLM Benchmark | — | 67% |
| EnigmaEval | — | 37.1% |
| Surface Evolver Bench | — | 93.1% |
Math Too close to call
Claude Opus 5: 86.2 (#8), GPT-5.6 Sol: 85.6 (#9)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| FrontierMath (Tiers 1-3) | 85.6% | 89.1% |
| FrontierMath Tier 4 | 73.2% | 82.9% |
| OTIS Mock AIME 2024-2025 | 98.9% | 100% |
| ProofBench | 99% | 83% |
| LMArena Math | 1531 | 1474 |
| FrontierMath Erdős | — | 0% |
Knowledge Claude Opus 5 leads
Claude Opus 5: 66.8 (#9), GPT-5.6 Sol: 64.3 (#18)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| GPQA Diamond | 93.9% | 93.5% |
| SimpleQA Verified | 59.9% | 69.7% |
| LMArena Expert | 1557 | 1516 |
| Vectara Hallucination Rate | — | 12.4% |
Multimodal Claude Opus 5 leads
Claude Opus 5: 50.8 (#8), GPT-5.6 Sol: 48.6 (#9)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| LMArena Vision | 1319 | 1281 |
| Blueprint-Bench 2 | 30.4% | 33.6% |
| Furniture Assembly | 60.8% | 56.7% |
| LMArena Document | 1516 | 1483 |
Multilingual Claude Opus 5 leads
Claude Opus 5: 58.8 (#4), GPT-5.6 Sol: 55.3 (#32)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| LMArena Non-English | 1501 | 1452 |
| LMArena Chinese | 1574 | 1527 |
| LMArena French | 1519 | 1477 |
| LMArena German | 1524 | 1476 |
| LMArena Japanese | 1516 | 1471 |
| LMArena Korean | 1521 | 1442 |
| LMArena Russian | 1507 | 1468 |
| LMArena Spanish | 1519 | 1441 |
Instruction Following Claude Opus 5 leads
Claude Opus 5: 79.2 (#7), GPT-5.6 Sol: 77.7 (#16)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| LMArena Instruction Following | 1517 | 1482 |
Long Context Claude Opus 5 leads
Claude Opus 5: 46.5 (#21), GPT-5.6 Sol: 45.4 (#42)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| LMArena Longer Query | 1515 | 1480 |
Writing & Preference Claude Opus 5 leads
Claude Opus 5: 79.2 (#1), GPT-5.6 Sol: 73.3 (#12)
| Benchmark | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| LMArena Text | 1507 | 1457 |
| LMArena Creative Writing | 1491 | 1448 |
| EQ-Bench Creative Writing | 2133 | 1972 |
| EQ-Bench 4 | 1385 | 1250 |
| LMArena Multi-Turn | 1499 | 1460 |
Frequently asked questions
Is Claude Opus 5 better than GPT-5.6 Sol?
Claude Opus 5 is the stronger model overall, scoring 67.8 to 65.0 on the Noometry Index.
Which is cheaper, Claude Opus 5 or GPT-5.6 Sol?
GPT-5.6 Sol is cheaper. It lists at $4 per million input tokens and $20 per million output tokens; Claude Opus 5 lists at $5 and $25.
Is Claude Opus 5 or GPT-5.6 Sol better for coding?
Claude Opus 5 scores higher on coding benchmarks: 67.5 versus 65.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Sol does, with 1.05M tokens against 1M.
How many benchmarks do Claude Opus 5 and GPT-5.6 Sol share?
57 benchmarks have published results for both models. Claude Opus 5 has 57 scored results on Noometry and GPT-5.6 Sol has 65.