Model comparison
Claude Opus 4.6 vs GPT-4.5
Claude Opus 4.6 is the stronger model overall, scoring 58.2 to 37.2 on the Noometry Index.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. Claude Opus 4.6 scores higher in 9 categories and GPT-4.5 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 4.6 leads 57.8 to 13.9.
- The biggest single-benchmark swing is ARC-AGI-1: 94% for Claude Opus 4.6 and 10.3% for GPT-4.5.
Side by side
| Claude Opus 4.6 | GPT-4.5 | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 58.2 | 37.2 |
| Released | 2026-02-04 | 2025-02-27 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | — |
| Max output | 128K | — |
| Input $ / M tokens | $5 | — |
| Output $ / M tokens | $25 | — |
| Results tracked | 68 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.6 leads
Claude Opus 4.6: 57.2 (#20), GPT-4.5: 42.2 (#109)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| WeirdML | 78% | 39.4% |
| LMArena Coding | 1536 | 1396 |
| SWE-bench Verified | 78.7% | — |
| FrontierCode | 26.6% | — |
| SWE-bench Verified (bash only) | 75.6% | — |
| Aider Polyglot | — | 44.9% |
| LMArena WebDev | 1547 | — |
| SWE-bench Multilingual | 72% | — |
| GSO | 41.2% | — |
| LiveBench Coding | — | 75.2% |
| ALE-Bench | 996.5 | — |
| AlgoTune | 1.47 | — |
Agentic & Tool Use Claude Opus 4.6 leads
Claude Opus 4.6: 51.1 (#4), GPT-4.5: 27.9 (#97)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| Cybench | 93% | 17.5% |
| Terminal-Bench | 79.8% | — |
| APEX-Agents | 46.3% | — |
| Remote Labor Index | 4.2% | — |
| τ²-bench Banking | 27.3% | — |
| DeepResearch Bench | 55.3% | — |
| GBAEval | 44.1% | — |
| LMArena Search | 1253 | — |
| METR Time Horizons | 78.9% | — |
| Vending-Bench 2 | 8,018 | — |
Reasoning Claude Opus 4.6 leads
Claude Opus 4.6: 57.8 (#23), GPT-4.5: 13.9 (#330)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| ARC-AGI-2 | 69.2% | 0.8% |
| SimpleBench | 67.6% | 34.5% |
| ARC-AGI-1 | 94% | 10.3% |
| EnigmaEval | 7.6% | 3.2% |
| LMArena Hard Prompts | 1527 | 1403 |
| Epoch Capabilities Index | 155.24 | 136.74 |
| ForecastBench | 60 | 61.7 |
| Kagi LLM Benchmark | 83.6% | — |
| NYT Connections (extended) | 92.1% | — |
| Chess Puzzles | 17% | — |
| Thematic Generalization | 80.6% | — |
| EBR-Bench | 12.7% | — |
| LiveBench Reasoning | — | 71.1% |
| Mystery Game Puzzles | 25% | — |
| DTBench | 91.2% | — |
| LiveBench Data Analysis | — | 64.3% |
| LMCA | 55.8% | — |
| LiveBench | — | 69% |
Math Claude Opus 4.6 leads
Claude Opus 4.6: 63.0 (#31), GPT-4.5: 32.6 (#211)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 94.4% | 37.8% |
| LMArena Math | 1519 | 1412 |
| FrontierMath (Tiers 1-3) | 66% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 78.5% | — |
| ProofBench | 50% | — |
| LiveBench Math | — | 69.3% |
| MATH Level 5 | — | 78.6% |
| FrontierMath (Feb 2025 set) | 40.7% | — |
| FrontierMath Tier 4 (v1) | 22.9% | — |
Knowledge Claude Opus 4.6 leads
Claude Opus 4.6: 61.9 (#26), GPT-4.5: 32.5 (#211)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| GPQA Diamond | 90.5% | 68.7% |
| Humanity's Last Exam | 34.4% | 5.4% |
| LMArena Expert | 1546 | 1394 |
| SimpleQA Verified | 47% | — |
| Confabulations | — | 13.6% |
| Vectara Hallucination Rate | 12.2% | — |
Multimodal Too close to call
Claude Opus 4.6: 37.3 (#74), GPT-4.5: 37.6 (#71)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| LMArena Vision | 1316 | 1195 |
| VPCT | — | 45% |
| Furniture Assembly | 28.3% | — |
| LMArena Document | 1507 | — |
Multilingual Claude Opus 4.6 leads
Claude Opus 4.6: 57.9 (#6), GPT-4.5: 52.5 (#83)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| LMArena Non-English | 1489 | 1413 |
| LMArena Chinese | 1551 | 1421 |
| LMArena French | 1513 | 1418 |
| LMArena German | 1502 | 1457 |
| LMArena Japanese | 1484 | 1416 |
| LMArena Korean | 1464 | 1392 |
| LMArena Russian | 1497 | 1419 |
| LMArena Spanish | 1510 | — |
Instruction Following Claude Opus 4.6 leads
Claude Opus 4.6: 79.5 (#4), GPT-4.5: 72.6 (#134)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| LMArena Instruction Following | 1523 | 1404 |
| LiveBench Instruction Following | — | 72.3% |
Long Context Claude Opus 4.6 leads
Claude Opus 4.6: 48.1 (#13), GPT-4.5: 40.4 (#155)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| LMArena Longer Query | 1520 | 1406 |
| Fiction.LiveBench | — | 63.9% |
| CL-bench | 20.7% | — |
| CL-bench Life | 17% | — |
Writing & Preference Claude Opus 4.6 leads
Claude Opus 4.6: 73.5 (#10), GPT-4.5: 56.9 (#134)
| Benchmark | Claude Opus 4.6 | GPT-4.5 |
|---|---|---|
| LMArena Text | 1503 | 1417 |
| LMArena Creative Writing | 1505 | 1394 |
| EQ-Bench Creative Writing | 1809 | 1258 |
| LMArena Multi-Turn | 1513 | 1444 |
| Short-Story Creative Writing | — | 75.6% |
| EQ-Bench 4 | 1223 | — |
| LiveBench Language | — | 61.5% |
Frequently asked questions
Is Claude Opus 4.6 better than GPT-4.5?
Claude Opus 4.6 is the stronger model overall, scoring 58.2 to 37.2 on the Noometry Index.
Is Claude Opus 4.6 or GPT-4.5 better for coding?
Claude Opus 4.6 scores higher on coding benchmarks: 57.2 versus 42.2 in the Noometry coding category.
How many benchmarks do Claude Opus 4.6 and GPT-4.5 share?
29 benchmarks have published results for both models. Claude Opus 4.6 has 68 scored results on Noometry and GPT-4.5 has 42.