Model comparison
GPT-4.1 vs GPT-4 Turbo
GPT-4.1 is the stronger model overall, scoring 35.7 to 30.3 on the Noometry Index.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. GPT-4.1 scores higher in 8 categories and GPT-4 Turbo in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-4.1 leads 21.4 to 8.6.
- The biggest single-benchmark swing is MATH Level 5: 83% for GPT-4.1 and 46.7% for GPT-4 Turbo.
- GPT-4.1 is cheaper at $2 / $8 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- GPT-4.1 accepts more context: 1.05M tokens versus 128K.
Side by side
| GPT-4.1 | GPT-4 Turbo | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 35.7 | 30.3 |
| Released | 2025-04-14 | 2023-11-06 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 128K |
| Max output | 33K | 4K |
| Input $ / M tokens | $2 | $10 |
| Output $ / M tokens | $8 | $30 |
| Results tracked | 52 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-4.1: 34.2 (#228), GPT-4 Turbo: 33.4 (#247)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| WeirdML | 39% | 18% |
| LMArena Coding | 1391 | 1268 |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| BigCodeBench Instruct | — | 48.2% |
| BigCodeBench Complete | — | 58.2% |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
| HumanEval+ | — | 86.6% |
| MBPP+ | — | 73.3% |
Agentic & Tool Use Not comparable
GPT-4.1: 34.4 (#40), GPT-4 Turbo: —
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
| METR Time Horizons | — | 36.7% |
Reasoning GPT-4 Turbo leads
GPT-4.1: 11.7 (#334), GPT-4 Turbo: 14.9 (#311)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| SimpleBench | 27% | 25.1% |
| Chess Puzzles | 6% | 6% |
| LMArena Hard Prompts | 1384 | 1251 |
| DTBench | 68.3% | 61.6% |
| LMCA | 25.6% | 9.8% |
| Epoch Capabilities Index | 136.79 | 127.25 |
| ForecastBench | 61.5 | 59.4 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 52.3% | — |
| ARC-AGI-1 | 5.5% | — |
| EnigmaEval | 2.2% | — |
Math GPT-4.1 leads
GPT-4.1: 21.4 (#269), GPT-4 Turbo: 8.6 (#316)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 0.7% |
| OTIS Mock AIME 2024-2025 | 38.3% | 6.7% |
| LMArena Math | 1369 | 1272 |
| MATH Level 5 | 83% | 46.7% |
| Omni-MATH | 47.1% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge GPT-4.1 leads
GPT-4.1: 37.0 (#152), GPT-4 Turbo: 24.3 (#263)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| GPQA Diamond | 66.9% | 46.6% |
| LMArena Expert | 1364 | 1223 |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| MMLU-Pro | 81.1% | — |
| Confabulations | — | 28.4% |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
| MMLU | — | 81.3% |
Multimodal GPT-4.1 leads
GPT-4.1: 37.5 (#66), GPT-4 Turbo: 30.0 (#104)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| LMArena Vision | 1210 | 1090 |
| GeoBench | 72% | — |
Multilingual GPT-4.1 leads
GPT-4.1: 49.6 (#128), GPT-4 Turbo: 40.7 (#211)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| LMArena Non-English | 1370 | 1245 |
| LMArena Chinese | 1382 | 1242 |
| LMArena French | 1383 | 1275 |
| LMArena German | 1380 | 1259 |
| LMArena Japanese | 1320 | 1194 |
| LMArena Korean | 1339 | 1187 |
| LMArena Russian | 1377 | 1259 |
| LMArena Spanish | 1375 | 1260 |
Instruction Following GPT-4.1 leads
GPT-4.1: 71.5 (#148), GPT-4 Turbo: 66.0 (#211)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| LMArena Instruction Following | 1367 | 1250 |
| IFEval | 83.8% | — |
Long Context GPT-4.1 leads
GPT-4.1: 40.0 (#158), GPT-4 Turbo: 38.0 (#200)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| LMArena Longer Query | 1385 | 1254 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-4.1 leads
GPT-4.1: 57.8 (#120), GPT-4 Turbo: 47.9 (#201)
| Benchmark | GPT-4.1 | GPT-4 Turbo |
|---|---|---|
| LMArena Text | 1383 | 1272 |
| LMArena Creative Writing | 1363 | 1269 |
| LMArena Multi-Turn | 1398 | 1267 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is GPT-4.1 better than GPT-4 Turbo?
GPT-4.1 is the stronger model overall, scoring 35.7 to 30.3 on the Noometry Index.
Which is cheaper, GPT-4.1 or GPT-4 Turbo?
GPT-4.1 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; GPT-4 Turbo lists at $10 and $30.
Is GPT-4.1 or GPT-4 Turbo better for coding?
They score almost the same on coding (34.2 vs 33.4); test both on your own repository before choosing.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 128K.
How many benchmarks do GPT-4.1 and GPT-4 Turbo share?
29 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and GPT-4 Turbo has 36.