Model comparison
Claude Sonnet 4 vs GPT-5 Mini
GPT-5 Mini is the stronger model overall, scoring 41.8 to 40.8 on the Noometry Index.
Last verified . 48 shared benchmarks.
Summary
- They share 48 benchmarks with published results for both. Claude Sonnet 4 scores higher in 3 categories and GPT-5 Mini in 7 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where GPT-5 Mini leads 35.6 to 26.2.
- The biggest single-benchmark swing is Fiction.LiveBench: 46.9% for Claude Sonnet 4 and 69.4% for GPT-5 Mini.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- GPT-5 Mini accepts more context: 400K tokens versus 200K.
Side by side
| Claude Sonnet 4 | GPT-5 Mini | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 40.8 | 41.8 |
| Released | 2025-05-22 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 400K |
| Max output | 64K | 128K |
| Input $ / M tokens | $3 | $0.25 |
| Output $ / M tokens | $15 | $2 |
| Results tracked | 58 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4 leads
Claude Sonnet 4: 43.5 (#88), GPT-5 Mini: 40.1 (#146)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| SWE-bench Verified (bash only) | 64.9% | 59.8% |
| SciCode | 40% | 39.2% |
| WeirdML | 46.1% | 52.7% |
| LMArena Coding | 1414 | 1406 |
| ALE-Bench | 655.35 | 799.77 |
| SWE-bench Verified | — | 64.7% |
| Aider Polyglot | 61.3% | — |
| SWE-bench Multilingual | — | 39.7% |
| GSO | 4.9% | — |
| AlgoTune | — | 1.38 |
Agentic & Tool Use Claude Sonnet 4 leads
Claude Sonnet 4: 38.5 (#31), GPT-5 Mini: 31.1 (#70)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| Terminal-Bench | — | 34.8% |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| METR Time Horizons | 62% | — |
| Vending-Bench 2 | — | -31.18 |
Reasoning GPT-5 Mini leads
Claude Sonnet 4: 22.9 (#187), GPT-5 Mini: 23.9 (#168)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| ARC-AGI-2 | 5.9% | 4.4% |
| Kagi LLM Benchmark | 73% | 70.3% |
| ARC-AGI-1 | 40% | 54.3% |
| CritPt | 0.3% | 0% |
| EnigmaEval | 3.1% | 8.2% |
| LMArena Hard Prompts | 1372 | 1380 |
| DTBench | 77.1% | 80.5% |
| LMCA | 29% | 34.2% |
| Epoch Capabilities Index | 141.69 | 145.52 |
| ForecastBench | 60.2 | 61 |
| SimpleBench | 45.5% | — |
| Chess Puzzles | — | 30% |
| Mystery Game Puzzles | — | 10% |
Math GPT-5 Mini leads
Claude Sonnet 4: 43.3 (#80), GPT-5 Mini: 46.7 (#69)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 86.7% |
| Omni-MATH | 60.2% | 72.2% |
| LMArena Math | 1375 | 1378 |
| MATH Level 5 | 84.4% | 97.8% |
| FrontierMath (Feb 2025 set) | 4.1% | 27.2% |
| FrontierMath Tier 4 (v1) | 0% | 6.3% |
| FrontierMath (Tiers 1-3) | — | 46.7% |
| FrontierMath Tier 4 | — | 12.2% |
| ProofBench | — | 9% |
Knowledge GPT-5 Mini leads
Claude Sonnet 4: 41.8 (#108), GPT-5 Mini: 45.6 (#86)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 79.2% | 75% |
| Humanity's Last Exam | 7.8% | 19.4% |
| MMLU-Pro | 84.3% | 83.5% |
| Confabulations | 13.2% | 13.3% |
| Vectara Hallucination Rate | 10.3% | 12.9% |
| GPQA (HELM) | 70.6% | 75.6% |
| LMArena Expert | 1372 | 1379 |
| SimpleQA Verified | — | 21.6% |
Multimodal GPT-5 Mini leads
Claude Sonnet 4: 26.2 (#121), GPT-5 Mini: 35.6 (#85)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1191 | 1202 |
| VPCT | 34% | 40.2% |
| GeoBench | 37% | — |
| MindCube | 44.8% | — |
Multilingual GPT-5 Mini leads
Claude Sonnet 4: 46.7 (#156), GPT-5 Mini: 48.9 (#137)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1333 | 1363 |
| LMArena Chinese | 1350 | 1385 |
| LMArena French | 1363 | 1386 |
| LMArena German | 1331 | 1366 |
| LMArena Japanese | 1302 | 1341 |
| LMArena Korean | 1291 | 1308 |
| LMArena Russian | 1355 | 1362 |
| LMArena Spanish | 1357 | 1355 |
Instruction Following GPT-5 Mini leads
Claude Sonnet 4: 71.7 (#145), GPT-5 Mini: 76.2 (#46)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| IFEval | 84% | 92.7% |
| LMArena Instruction Following | 1376 | 1357 |
Long Context GPT-5 Mini leads
Claude Sonnet 4: 33.7 (#259), GPT-5 Mini: 41.9 (#132)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| Fiction.LiveBench | 46.9% | 69.4% |
| LMArena Longer Query | 1398 | 1355 |
Writing & Preference Claude Sonnet 4 leads
Claude Sonnet 4: 57.1 (#132), GPT-5 Mini: 55.2 (#148)
| Benchmark | Claude Sonnet 4 | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1351 | 1373 |
| LMArena Creative Writing | 1345 | 1325 |
| Short-Story Creative Writing | 81.4% | 83.1% |
| EQ-Bench Creative Writing | 1483 | 1313 |
| WildBench | 83.8% | 85.5% |
| LMArena Multi-Turn | 1376 | 1363 |
Frequently asked questions
Is Claude Sonnet 4 better than GPT-5 Mini?
GPT-5 Mini is the stronger model overall, scoring 41.8 to 40.8 on the Noometry Index.
Which is cheaper, Claude Sonnet 4 or GPT-5 Mini?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; Claude Sonnet 4 lists at $3 and $15.
Is Claude Sonnet 4 or GPT-5 Mini better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 43.5 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 200K.
How many benchmarks do Claude Sonnet 4 and GPT-5 Mini share?
48 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and GPT-5 Mini has 60.