Model comparison
Claude Sonnet 5 vs GPT-5.4 Pro
GPT-5.4 Pro is the stronger model overall, scoring 58.8 to 54.5 on the Noometry Index. Claude Sonnet 5 costs 17× less per token, which makes it the better buy when GPT-5.4 Pro's lead doesn't matter for your workload.
Last verified . 9 shared benchmarks.
Summary
- They share 9 benchmarks with published results for both. Claude Sonnet 5 scores higher in 1 category and GPT-5.4 Pro in 3 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.4 Pro leads 71.0 to 48.2.
- The biggest single-benchmark swing is FrontierMath Tier 4: 29.3% for Claude Sonnet 5 and 58.5% for GPT-5.4 Pro.
- Claude Sonnet 5 is cheaper at $2 / $10 per million input/output tokens, against $30 / $180 for GPT-5.4 Pro.
- GPT-5.4 Pro accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Sonnet 5 | GPT-5.4 Pro | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 54.5 | 58.8 |
| Released | 2026-06-29 | 2026-03-05 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $2 | $30 |
| Output $ / M tokens | $10 | $180 |
| Results tracked | 51 | 16 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 5 leads
Claude Sonnet 5: 55.5 (#20), GPT-5.4 Pro: 42.6 (#89)
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| WeirdML | 68.8% | 57.4% |
| DeepSWE | 53.8% | — |
| FrontierCode | 42.7% | — |
| CursorBench | 34.1% | — |
| LMArena WebDev | 1540 | — |
| SciCode | 53.6% | — |
| GSO | 37.3% | — |
| LMArena Coding | 1483 | — |
| ALE-Bench | 1,463 | — |
Agentic & Tool Use Not comparable
Claude Sonnet 5: 42.7 (#16), GPT-5.4 Pro: —
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| APEX-Agents | 54.5% | — |
| GBAEval | 65.3% | — |
| LMArena Search | 1194 | — |
| Vending-Bench 2 | 6,378 | — |
Reasoning GPT-5.4 Pro leads
Claude Sonnet 5: 48.2 (#37), GPT-5.4 Pro: 71.0 (#10)
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| SimpleBench | 60.6% | 74.1% |
| CritPt | 16.9% | 30% |
| Chess Puzzles | 35% | 58.6% |
| Epoch Capabilities Index | 156.34 | 159.12 |
| ARC-AGI-2 | — | 83.3% |
| NYT Connections (extended) | 75.1% | — |
| ARC-AGI-1 | — | 94.5% |
| EnigmaEval | — | 23.8% |
| LMArena Hard Prompts | 1461 | — |
| Mystery Game Puzzles | 35% | — |
| DTBench | 92.5% | — |
| LMCA | 50% | — |
| Surface Evolver Bench | 60% | — |
| Bench to the Future 3 | 0.14 | — |
| ForecastBench | 61.1 | — |
Math GPT-5.4 Pro leads
Claude Sonnet 5: 66.1 (#21), GPT-5.4 Pro: 72.0 (#17)
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| FrontierMath (Tiers 1-3) | 65.6% | 82.5% |
| FrontierMath Tier 4 | 29.3% | 58.5% |
| OTIS Mock AIME 2024-2025 | 94.7% | — |
| ProofBench | 77% | — |
| LMArena Math | 1467 | — |
| FrontierMath (Feb 2025 set) | — | 50% |
| FrontierMath Tier 4 (v1) | — | 37.5% |
Knowledge GPT-5.4 Pro leads
Claude Sonnet 5: 55.7 (#42), GPT-5.4 Pro: 68.3 (#6)
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| GPQA Diamond | 90.5% | 94.6% |
| SimpleQA Verified | 33.7% | 46.3% |
| Humanity's Last Exam | — | 44.3% |
| Vectara Hallucination Rate | — | 8.3% |
| LMArena Expert | 1491 | — |
Multimodal Not comparable
Claude Sonnet 5: 42.4 (#26), GPT-5.4 Pro: —
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| LMArena Vision | 1275 | — |
| Blueprint-Bench 2 | 24.9% | — |
| LMArena Document | 1466 | — |
Multilingual Not comparable
Claude Sonnet 5: 54.0 (#55), GPT-5.4 Pro: —
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| LMArena Non-English | 1431 | — |
| LMArena Chinese | 1475 | — |
| LMArena French | 1457 | — |
| LMArena German | 1448 | — |
| LMArena Japanese | 1422 | — |
| LMArena Korean | 1411 | — |
| LMArena Russian | 1452 | — |
| LMArena Spanish | 1436 | — |
Instruction Following Not comparable
Claude Sonnet 5: 76.6 (#39), GPT-5.4 Pro: —
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| LMArena Instruction Following | 1453 | — |
Long Context Not comparable
Claude Sonnet 5: 44.8 (#53), GPT-5.4 Pro: —
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| LMArena Longer Query | 1464 | — |
Writing & Preference Not comparable
Claude Sonnet 5: 69.4 (#24), GPT-5.4 Pro: —
| Benchmark | Claude Sonnet 5 | GPT-5.4 Pro |
|---|---|---|
| LMArena Text | 1443 | — |
| LMArena Creative Writing | 1416 | — |
| EQ-Bench Creative Writing | 1794 | — |
| EQ-Bench 4 | 1236 | — |
| LMArena Multi-Turn | 1454 | — |
Frequently asked questions
Is Claude Sonnet 5 better than GPT-5.4 Pro?
GPT-5.4 Pro is the stronger model overall, scoring 58.8 to 54.5 on the Noometry Index. Claude Sonnet 5 costs 17× less per token, which makes it the better buy when GPT-5.4 Pro's lead doesn't matter for your workload.
Which is cheaper, Claude Sonnet 5 or GPT-5.4 Pro?
Claude Sonnet 5 is cheaper. It lists at $2 per million input tokens and $10 per million output tokens; GPT-5.4 Pro lists at $30 and $180.
Is Claude Sonnet 5 or GPT-5.4 Pro better for coding?
Claude Sonnet 5 scores higher on coding benchmarks: 55.5 versus 42.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 Pro does, with 1.05M tokens against 1M.
How many benchmarks do Claude Sonnet 5 and GPT-5.4 Pro share?
9 benchmarks have published results for both models. Claude Sonnet 5 has 51 scored results on Noometry and GPT-5.4 Pro has 16.