Model comparison
Claude Opus 4.5 vs GPT-5.4
GPT-5.4 is the stronger model overall, scoring 59.4 to 50.5 on the Noometry Index.
Last verified . 55 shared benchmarks.
Summary
- They share 55 benchmarks with published results for both. Claude Opus 4.5 scores higher in 3 categories and GPT-5.4 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 leads 73.5 to 38.6.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 34.4% for Claude Opus 4.5 and 78.6% for GPT-5.4.
- GPT-5.4 is cheaper at $2.50 / $15 per million input/output tokens, against $5 / $25 for Claude Opus 4.5.
- GPT-5.4 accepts more context: 1.05M tokens versus 200K.
Side by side
| Claude Opus 4.5 | GPT-5.4 | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 50.5 | 59.4 |
| Released | 2025-11-01 | 2026-03-05 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1.05M |
| Max output | 64K | 128K |
| Input $ / M tokens | $5 | $2.50 |
| Output $ / M tokens | $25 | $15 |
| Results tracked | 69 | 68 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.5 leads
Claude Opus 4.5: 54.8 (#27), GPT-5.4: 52.6 (#33)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| SWE-bench Verified | 76.7% | 76.9% |
| LMArena WebDev | 1494 | 1465 |
| GSO | 26.5% | 31.4% |
| WeirdML | 63.7% | 77.7% |
| LMArena Coding | 1504 | 1497 |
| ALE-Bench | 1,025 | 1,607 |
| AlgoTune | 1.77 | 1.85 |
| DeepSWE | — | 51.8% |
| SWE-bench Verified (bash only) | 76.8% | — |
| SWE-bench Multilingual | 70.7% | — |
| SciCode | — | 56.6% |
| MirrorCode | — | 15.6% |
Agentic & Tool Use Too close to call
Claude Opus 4.5: 47.3 (#12), GPT-5.4: 46.5 (#13)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| Terminal-Bench | 63.1% | 81.8% |
| τ²-bench Banking | 24.7% | 39.4% |
| DeepResearch Bench | 54.8% | 35.1% |
| LMArena Search | 1180 | 1197 |
| METR Time Horizons | 75% | 74.3% |
| Vending-Bench 2 | 4,967 | 6,144 |
| APEX-Agents | — | 52.4% |
| Berkeley Function Calling Leaderboard | 77.5% | — |
| GDPval | 45.5% | — |
| Remote Labor Index | 3.8% | — |
| τ²-bench Airline | 84% | — |
| τ²-bench Retail | 79.6% | — |
| τ²-bench Telecom | 92.3% | — |
| Cybench | 82% | — |
| OSWorld | 66.3% | — |
| PostTrainBench | — | 19% |
| BALROG | 43.5% | — |
| GBAEval | — | 45.1% |
Reasoning GPT-5.4 leads
Claude Opus 4.5: 42.6 (#51), GPT-5.4: 61.8 (#19)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| ARC-AGI-2 | 37.6% | 74% |
| Kagi LLM Benchmark | 80.2% | 63.8% |
| NYT Connections (extended) | 52.5% | 91.3% |
| ARC-AGI-1 | 80% | 93.7% |
| Chess Puzzles | 12% | 44% |
| EnigmaEval | 11.9% | 16% |
| EBR-Bench | 14.3% | 25.4% |
| LMArena Hard Prompts | 1476 | 1485 |
| Mystery Game Puzzles | 22% | 37% |
| DTBench | 89.9% | 94.4% |
| LMCA | 44.5% | 52% |
| Epoch Capabilities Index | 150.09 | 156.81 |
| ForecastBench | 60.7 | 59.5 |
| SimpleBench | 62% | — |
| CritPt | — | 23.4% |
| Thematic Generalization | — | 80% |
Math GPT-5.4 leads
Claude Opus 4.5: 38.6 (#132), GPT-5.4: 73.5 (#19)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 34.4% | 78.6% |
| FrontierMath Tier 4 | 4.9% | 49% |
| OTIS Mock AIME 2024-2025 | 86.1% | 97.8% |
| ProofBench | 36% | 56% |
| LMArena Math | 1463 | 1488 |
| FrontierMath (Feb 2025 set) | 20.7% | 47.6% |
| FrontierMath Tier 4 (v1) | 4.2% | 27.1% |
| MathArena Final-Answer Competitions | — | 83.1% |
Knowledge GPT-5.4 leads
Claude Opus 4.5: 56.5 (#44), GPT-5.4: 65.3 (#14)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| GPQA Diamond | 86% | 93.3% |
| Humanity's Last Exam | 25.2% | 36.2% |
| SimpleQA Verified | 45.7% | 45.1% |
| Vectara Hallucination Rate | 10.9% | 7% |
| LMArena Expert | 1487 | 1507 |
Multimodal GPT-5.4 leads
Claude Opus 4.5: 31.4 (#107), GPT-5.4: 43.7 (#20)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| Furniture Assembly | 28.3% | 37.5% |
| LMArena Document | 1462 | 1471 |
| LMArena Vision | — | 1303 |
| GeoBench | 75% | — |
| VPCT | 40% | — |
| Blueprint-Bench 2 | — | 27.1% |
Multilingual GPT-5.4 leads
Claude Opus 4.5: 54.3 (#47), GPT-5.4: 56.2 (#23)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| LMArena Non-English | 1438 | 1465 |
| LMArena Chinese | 1470 | 1519 |
| LMArena French | 1471 | 1493 |
| LMArena German | 1449 | 1472 |
| LMArena Japanese | 1416 | 1485 |
| LMArena Korean | 1424 | 1448 |
| LMArena Russian | 1447 | 1480 |
| LMArena Spanish | 1458 | 1454 |
Instruction Following Too close to call
Claude Opus 4.5: 77.5 (#19), GPT-5.4: 77.1 (#27)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| LMArena Instruction Following | 1478 | 1469 |
Long Context GPT-5.4 leads
Claude Opus 4.5: 46.5 (#22), GPT-5.4: 50.3 (#8)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| CL-bench | 21.1% | 27.9% |
| LMArena Longer Query | 1480 | 1473 |
| CL-bench Life | — | 21.7% |
Writing & Preference GPT-5.4 leads
Claude Opus 4.5: 68.1 (#28), GPT-5.4: 71.9 (#17)
| Benchmark | Claude Opus 4.5 | GPT-5.4 |
|---|---|---|
| LMArena Text | 1451 | 1469 |
| LMArena Creative Writing | 1445 | 1439 |
| EQ-Bench Creative Writing | 1687 | 1840 |
| LMArena Multi-Turn | 1466 | 1482 |
| EQ-Bench 4 | — | 1272 |
Frequently asked questions
Is Claude Opus 4.5 better than GPT-5.4?
GPT-5.4 is the stronger model overall, scoring 59.4 to 50.5 on the Noometry Index.
Which is cheaper, Claude Opus 4.5 or GPT-5.4?
GPT-5.4 is cheaper. It lists at $2.50 per million input tokens and $15 per million output tokens; Claude Opus 4.5 lists at $5 and $25.
Is Claude Opus 4.5 or GPT-5.4 better for coding?
Claude Opus 4.5 scores higher on coding benchmarks: 54.8 versus 52.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 does, with 1.05M tokens against 200K.
How many benchmarks do Claude Opus 4.5 and GPT-5.4 share?
55 benchmarks have published results for both models. Claude Opus 4.5 has 69 scored results on Noometry and GPT-5.4 has 68.