Model comparison
Claude Opus 4.8 vs GPT-5.6 Sol
GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 60.7 on the Noometry Index.
Last verified . 59 shared benchmarks.
Summary
- They share 59 benchmarks with published results for both. Claude Opus 4.8 scores higher in 1 category and GPT-5.6 Sol in 9 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Sol leads 74.8 to 64.7.
- The biggest single-benchmark swing is Chess Puzzles: 34% for Claude Opus 4.8 and 64% 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 4.8.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Opus 4.8 | GPT-5.6 Sol | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 60.7 | 65.0 |
| Released | 2026-05-28 | 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 | 65 | 65 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Sol leads
Claude Opus 4.8: 59.9 (#12), GPT-5.6 Sol: 65.1 (#7)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| DeepSWE | 59% | 72.7% |
| FrontierCode | 46.5% | 47.5% |
| LMArena WebDev | 1556 | 1618 |
| SciCode | 53.5% | 57.1% |
| GSO | 47.1% | 76.5% |
| WeirdML | 82.9% | 89.4% |
| LMArena Coding | 1490 | 1498 |
| ALE-Bench | 1,564 | 2,177 |
| CursorBench | — | 41.7% |
| FrontierSWE | — | 32.2% |
| MirrorCode | — | 20% |
Agentic & Tool Use GPT-5.6 Sol leads
Claude Opus 4.8: 47.6 (#11), GPT-5.6 Sol: 50.3 (#7)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| APEX-Agents | 48.9% | 51.4% |
| OSWorld 2.0 | 20.6% | 27.3% |
| τ²-bench Banking | 39.7% | 46.9% |
| PostTrainBench | 33.8% | 36.2% |
| GBAEval | 70.9% | 52.6% |
| GDP.pdf | 24% | 30.7% |
| LMArena Search | 1204 | 1257 |
| Vending-Bench 2 | 5,787 | 9,619 |
| Remote Labor Index | 8.3% | — |
| DeepResearch Bench | 50.2% | — |
| BALROG | — | 60% |
Reasoning GPT-5.6 Sol leads
Claude Opus 4.8: 64.7 (#16), GPT-5.6 Sol: 74.8 (#8)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| ARC-AGI-2 | 72.1% | 92.5% |
| SimpleBench | 64.8% | 71.7% |
| Kagi LLM Benchmark | 88.8% | 67% |
| NYT Connections (extended) | 91.1% | 93.8% |
| ARC-AGI-1 | 92.5% | 97.5% |
| CritPt | 20.9% | 32.3% |
| Chess Puzzles | 34% | 64% |
| EnigmaEval | 23.5% | 37.1% |
| EBR-Bench | 28.6% | 44.8% |
| LMArena Hard Prompts | 1482 | 1484 |
| Mystery Game Puzzles | 36% | 58% |
| DTBench | 94.9% | 96% |
| LMCA | 57.5% | 59.2% |
| Surface Evolver Bench | 87.5% | 93.1% |
| Bench to the Future 3 | 0.14 | 0.14 |
| Epoch Capabilities Index | 158.21 | 161.66 |
| ForecastBench | 59.9 | — |
Math GPT-5.6 Sol leads
Claude Opus 4.8: 78.4 (#13), GPT-5.6 Sol: 85.6 (#9)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| FrontierMath (Tiers 1-3) | 80% | 89.1% |
| FrontierMath Tier 4 | 56.1% | 82.9% |
| OTIS Mock AIME 2024-2025 | 98.3% | 100% |
| ProofBench | 69% | 83% |
| LMArena Math | 1487 | 1474 |
| MathArena Final-Answer Competitions | 91.8% | — |
| FrontierMath (Feb 2025 set) | 47.2% | — |
| FrontierMath Erdős | — | 0% |
| FrontierMath Tier 4 (v1) | 31.3% | — |
Knowledge GPT-5.6 Sol leads
Claude Opus 4.8: 61.3 (#29), GPT-5.6 Sol: 64.3 (#18)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| GPQA Diamond | 91% | 93.5% |
| SimpleQA Verified | 53% | 69.7% |
| LMArena Expert | 1502 | 1516 |
| Vectara Hallucination Rate | — | 12.4% |
Multimodal GPT-5.6 Sol leads
Claude Opus 4.8: 42.9 (#26), GPT-5.6 Sol: 48.6 (#9)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| LMArena Vision | 1294 | 1281 |
| Blueprint-Bench 2 | 14.5% | 33.6% |
| Furniture Assembly | 42.5% | 56.7% |
| LMArena Document | 1475 | 1483 |
Multilingual Too close to call
Claude Opus 4.8: 55.2 (#33), GPT-5.6 Sol: 55.3 (#32)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| LMArena Non-English | 1450 | 1452 |
| LMArena Chinese | 1507 | 1527 |
| LMArena French | 1481 | 1477 |
| LMArena German | 1472 | 1476 |
| LMArena Japanese | 1440 | 1471 |
| LMArena Korean | 1432 | 1442 |
| LMArena Russian | 1474 | 1468 |
| LMArena Spanish | 1466 | 1441 |
Instruction Following Too close to call
Claude Opus 4.8: 77.4 (#24), GPT-5.6 Sol: 77.7 (#16)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| LMArena Instruction Following | 1476 | 1482 |
Long Context Too close to call
Claude Opus 4.8: 45.4 (#35), GPT-5.6 Sol: 45.4 (#42)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| LMArena Longer Query | 1483 | 1480 |
Writing & Preference GPT-5.6 Sol leads
Claude Opus 4.8: 72.0 (#16), GPT-5.6 Sol: 73.3 (#12)
| Benchmark | Claude Opus 4.8 | GPT-5.6 Sol |
|---|---|---|
| LMArena Text | 1461 | 1457 |
| LMArena Creative Writing | 1454 | 1448 |
| EQ-Bench Creative Writing | 1840 | 1972 |
| EQ-Bench 4 | 1281 | 1250 |
| LMArena Multi-Turn | 1476 | 1460 |
Frequently asked questions
Is Claude Opus 4.8 better than GPT-5.6 Sol?
GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 60.7 on the Noometry Index.
Which is cheaper, Claude Opus 4.8 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 4.8 lists at $5 and $25.
Is Claude Opus 4.8 or GPT-5.6 Sol better for coding?
GPT-5.6 Sol scores higher on coding benchmarks: 65.1 versus 59.9 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 4.8 and GPT-5.6 Sol share?
59 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and GPT-5.6 Sol has 65.