Model comparison
Claude Opus 4 vs GPT-6 Astra
GPT-6 Astra is the stronger model overall, scoring 70.3 to 42.9 on the Noometry Index.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. Claude Opus 4 scores higher in 1 category and GPT-6 Astra in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 27.4.
- The biggest single-benchmark swing is ARC-AGI-2: 8.6% for Claude Opus 4 and 95% for GPT-6 Astra.
- GPT-6 Astra is cheaper at $10 / $50 per million input/output tokens, against $15 / $75 for Claude Opus 4.
- GPT-6 Astra accepts more context: 1.05M tokens versus 200K.
Side by side
| Claude Opus 4 | GPT-6 Astra | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 42.9 | 70.3 |
| Released | 2025-05-22 | 2026-09-03 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1.05M |
| Max output | 32K | 128K |
| Input $ / M tokens | $15 | $10 |
| Output $ / M tokens | $75 | $50 |
| Results tracked | 56 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
Claude Opus 4: 47.8 (#55), GPT-6 Astra: 70.5 (#3)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| WeirdML | 43.7% | 93.6% |
| LMArena Coding | 1442 | 1488 |
| SWE-bench Verified | 70.7% | — |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| SWE-bench Verified (bash only) | 67.6% | — |
| Aider Polyglot | 72% | — |
| LMArena WebDev | — | 1792 |
| FrontierSWE | — | 65.5% |
| SciCode | — | 56.5% |
| GSO | 6.9% | — |
| MirrorCode | — | 46.7% |
| ALE-Bench | — | 2,951 |
| AlgoTune | 1.33 | — |
Agentic & Tool Use GPT-6 Astra leads
Claude Opus 4: 34.5 (#39), GPT-6 Astra: 52.3 (#3)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| APEX-Agents | — | 64.7% |
| Remote Labor Index | — | 20.8% |
| Cybench | 38% | — |
| DeepResearch Bench | 46.8% | — |
| BALROG | — | 68.3% |
| LMArena Search | 1127 | — |
| METR Time Horizons | 63.9% | — |
| Vending-Bench 2 | — | 15,515 |
Reasoning GPT-6 Astra leads
Claude Opus 4: 27.4 (#104), GPT-6 Astra: 85.1 (#1)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| ARC-AGI-2 | 8.6% | 95% |
| ARC-AGI-1 | 35.7% | 98.5% |
| CritPt | 0.3% | 31.7% |
| LMArena Hard Prompts | 1399 | 1464 |
| DTBench | 81.6% | 97.3% |
| LMCA | 37.4% | 64.4% |
| Epoch Capabilities Index | 142.67 | 166.6 |
| SimpleBench | 58.8% | — |
| Kagi LLM Benchmark | 74.3% | — |
| NYT Connections (extended) | — | 98.1% |
| Chess Puzzles | — | 72% |
| EnigmaEval | 5.6% | — |
| EBR-Bench | — | 76.2% |
| Mystery Game Puzzles | — | 84% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | 61.1 | — |
Math GPT-6 Astra leads
Claude Opus 4: 40.2 (#80), GPT-6 Astra: 93.4 (#1)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 100% |
| LMArena Math | 1391 | 1463 |
| FrontierMath (Tiers 1-3) | — | 93.7% |
| FrontierMath Tier 4 | — | 97.6% |
| ProofBench | — | 99% |
| 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 GPT-6 Astra leads
Claude Opus 4: 43.9 (#82), GPT-6 Astra: 75.3 (#1)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| GPQA Diamond | 76.3% | 95.8% |
| Humanity's Last Exam | 10.7% | 54.8% |
| Vectara Hallucination Rate | 12% | 8.7% |
| LMArena Expert | 1384 | 1486 |
| SimpleQA Verified | — | 75.6% |
| MMLU-Pro | 87.5% | — |
| Confabulations | 15.9% | — |
| GPQA (HELM) | 70.8% | — |
Multimodal GPT-6 Astra leads
Claude Opus 4: 30.9 (#101), GPT-6 Astra: 55.1 (#3)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| LMArena Vision | 1192 | 1280 |
| GeoBench | 49% | — |
| VPCT | 38% | — |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |
Multilingual GPT-6 Astra leads
Claude Opus 4: 49.0 (#133), GPT-6 Astra: 54.0 (#53)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1362 | 1431 |
| LMArena Chinese | 1386 | 1495 |
| LMArena French | 1370 | 1467 |
| LMArena Russian | 1392 | 1434 |
| LMArena Spanish | 1388 | 1412 |
| LMArena German | 1391 | — |
| LMArena Japanese | 1332 | — |
| LMArena Korean | 1318 | — |
Instruction Following Too close to call
Claude Opus 4: 77.3 (#26), GPT-6 Astra: 76.6 (#37)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1406 | 1454 |
| IFEval | 91.8% | — |
Long Context GPT-6 Astra leads
Claude Opus 4: 39.5 (#167), GPT-6 Astra: 44.7 (#57)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1422 | 1460 |
| Fiction.LiveBench | 61.1% | — |
Writing & Preference GPT-6 Astra leads
Claude Opus 4: 61.4 (#85), GPT-6 Astra: 75.7 (#7)
| Benchmark | Claude Opus 4 | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1377 | 1443 |
| LMArena Creative Writing | 1386 | 1423 |
| EQ-Bench Creative Writing | 1580 | 2173 |
| LMArena Multi-Turn | 1396 | 1457 |
| Short-Story Creative Writing | 83.6% | — |
| WildBench | 85.2% | — |
Frequently asked questions
Is Claude Opus 4 better than GPT-6 Astra?
GPT-6 Astra is the stronger model overall, scoring 70.3 to 42.9 on the Noometry Index.
Which is cheaper, Claude Opus 4 or GPT-6 Astra?
GPT-6 Astra is cheaper. It lists at $10 per million input tokens and $50 per million output tokens; Claude Opus 4 lists at $15 and $75.
Is Claude Opus 4 or GPT-6 Astra better for coding?
GPT-6 Astra scores higher on coding benchmarks: 70.5 versus 47.8 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Astra does, with 1.05M tokens against 200K.
How many benchmarks do Claude Opus 4 and GPT-6 Astra share?
27 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and GPT-6 Astra has 51.