Model comparison
Claude Sonnet 4 vs GPT-4 Turbo
Claude Sonnet 4 is the stronger model overall, scoring 40.8 to 30.5 on the Noometry Index.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. Claude Sonnet 4 scores higher in 7 categories and GPT-4 Turbo in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Sonnet 4 leads 43.3 to 9.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 71.1% for Claude Sonnet 4 and 6.7% for GPT-4 Turbo.
- Claude Sonnet 4 is cheaper at $3 / $15 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- Claude Sonnet 4 accepts more context: 200K tokens versus 128K.
Side by side
| Claude Sonnet 4 | GPT-4 Turbo | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 40.8 | 30.5 |
| Released | 2025-05-22 | 2023-11-06 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 128K |
| Max output | 64K | 4K |
| Input $ / M tokens | $3 | $10 |
| Output $ / M tokens | $15 | $30 |
| Results tracked | 58 | 36 |
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-4 Turbo: 33.8 (#249)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| WeirdML | 46.1% | 18% |
| LMArena Coding | 1414 | 1268 |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| SciCode | 40% | — |
| GSO | 4.9% | — |
| BigCodeBench Instruct | — | 48.2% |
| BigCodeBench Complete | — | 58.2% |
| ALE-Bench | 655.35 | — |
| HumanEval+ | — | 86.6% |
| MBPP+ | — | 73.3% |
Agentic & Tool Use Not comparable
Claude Sonnet 4: 38.5 (#31), GPT-4 Turbo: —
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| METR Time Horizons | 62% | 36.7% |
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
Reasoning Claude Sonnet 4 leads
Claude Sonnet 4: 22.9 (#187), GPT-4 Turbo: 15.3 (#317)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| SimpleBench | 45.5% | 25.1% |
| LMArena Hard Prompts | 1372 | 1251 |
| DTBench | 77.1% | 61.6% |
| LMCA | 29% | 9.8% |
| Epoch Capabilities Index | 141.69 | 127.25 |
| ForecastBench | 60.2 | 59.4 |
| ARC-AGI-2 | 5.9% | — |
| Kagi LLM Benchmark | 73% | — |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| Chess Puzzles | — | 6% |
| EnigmaEval | 3.1% | — |
Math Claude Sonnet 4 leads
Claude Sonnet 4: 43.3 (#80), GPT-4 Turbo: 9.0 (#322)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 6.7% |
| LMArena Math | 1375 | 1272 |
| MATH Level 5 | 84.4% | 46.7% |
| FrontierMath (Tiers 1-3) | — | 0.7% |
| Omni-MATH | 60.2% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Claude Sonnet 4 leads
Claude Sonnet 4: 41.8 (#108), GPT-4 Turbo: 24.3 (#268)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| GPQA Diamond | 79.2% | 46.6% |
| Confabulations | 13.2% | 28.4% |
| LMArena Expert | 1372 | 1223 |
| Humanity's Last Exam | 7.8% | — |
| MMLU-Pro | 84.3% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |
| MMLU | — | 81.3% |
Multimodal GPT-4 Turbo leads
Claude Sonnet 4: 26.2 (#121), GPT-4 Turbo: 30.6 (#110)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| LMArena Vision | 1191 | 1090 |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| MindCube | 44.8% | — |
Multilingual Claude Sonnet 4 leads
Claude Sonnet 4: 46.7 (#156), GPT-4 Turbo: 40.5 (#216)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| LMArena Non-English | 1333 | 1245 |
| LMArena Chinese | 1350 | 1242 |
| LMArena French | 1363 | 1276 |
| LMArena German | 1331 | 1259 |
| LMArena Japanese | 1302 | 1194 |
| LMArena Korean | 1291 | 1187 |
| LMArena Russian | 1355 | 1259 |
| LMArena Spanish | 1357 | 1260 |
Instruction Following Claude Sonnet 4 leads
Claude Sonnet 4: 71.7 (#145), GPT-4 Turbo: 65.8 (#216)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| LMArena Instruction Following | 1376 | 1249 |
| IFEval | 84% | — |
Long Context GPT-4 Turbo leads
Claude Sonnet 4: 33.7 (#259), GPT-4 Turbo: 38.0 (#206)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| LMArena Longer Query | 1398 | 1254 |
| Fiction.LiveBench | 46.9% | — |
Writing & Preference Claude Sonnet 4 leads
Claude Sonnet 4: 57.1 (#132), GPT-4 Turbo: 47.7 (#206)
| Benchmark | Claude Sonnet 4 | GPT-4 Turbo |
|---|---|---|
| LMArena Text | 1351 | 1272 |
| LMArena Creative Writing | 1345 | 1269 |
| LMArena Multi-Turn | 1376 | 1267 |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
| WildBench | 83.8% | — |
Frequently asked questions
Is Claude Sonnet 4 better than GPT-4 Turbo?
Claude Sonnet 4 is the stronger model overall, scoring 40.8 to 30.5 on the Noometry Index.
Which is cheaper, Claude Sonnet 4 or GPT-4 Turbo?
Claude Sonnet 4 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; GPT-4 Turbo lists at $10 and $30.
Is Claude Sonnet 4 or GPT-4 Turbo better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 43.5 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
Claude Sonnet 4 does, with 200K tokens against 128K.
How many benchmarks do Claude Sonnet 4 and GPT-4 Turbo share?
29 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and GPT-4 Turbo has 36.