Model comparison
GPT-4 vs GPT-4.1
GPT-4.1 is the stronger model overall, scoring 35.9 to 29.1 on the Noometry Index.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. GPT-4 scores higher in 1 category and GPT-4.1 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-4.1 leads 57.6 to 34.9.
- The biggest single-benchmark swing is MATH Level 5: 23% for GPT-4 and 83% for GPT-4.1.
- GPT-4.1 is cheaper at $2 / $8 per million input/output tokens, against $30 / $60 for GPT-4.
- GPT-4.1 accepts more context: 1.05M tokens versus 8K.
Side by side
| GPT-4 | GPT-4.1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 29.1 | 35.9 |
| Released | 2023-03-14 | 2025-04-14 |
| Weights | Proprietary | Proprietary |
| Context window | 8K | 1.05M |
| Max output | 8K | 33K |
| Input $ / M tokens | $30 | $2 |
| Output $ / M tokens | $60 | $8 |
| Results tracked | 38 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4.1 leads
GPT-4: 31.6 (#283), GPT-4.1: 34.4 (#238)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| WeirdML | 12.4% | 39% |
| LMArena Coding | 1254 | 1391 |
| SWE-bench Verified | — | 48.5% |
| SWE-bench Verified (bash only) | — | 39.6% |
| Aider Polyglot | — | 52.4% |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| CadEval | — | 42% |
| ALE-Bench | — | 558.1 |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, GPT-4.1: 34.7 (#43)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 54% |
| METR Time Horizons | 36.1% | — |
Reasoning GPT-4 leads
GPT-4: 17.8 (#289), GPT-4.1: 11.7 (#339)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| Chess Puzzles | 4% | 6% |
| LMArena Hard Prompts | 1241 | 1384 |
| DTBench | 62.7% | 68.3% |
| LMCA | 17.1% | 25.6% |
| Epoch Capabilities Index | 125.89 | 136.78 |
| ForecastBench | 57.8 | 61.5 |
| ARC-AGI-2 | — | 0.4% |
| SimpleBench | — | 27% |
| Kagi LLM Benchmark | — | 52.3% |
| ARC-AGI-1 | — | 5.5% |
| EnigmaEval | — | 2.2% |
| Mystery Game Puzzles | 12% | — |
| BIG-Bench Hard | 75.1% | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math GPT-4.1 leads
GPT-4: 10.8 (#309), GPT-4.1: 22.3 (#280)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 38.3% |
| LMArena Math | 1269 | 1370 |
| MATH Level 5 | 23% | 83% |
| FrontierMath (Tiers 1-3) | — | 6% |
| Omni-MATH | — | 47.1% |
| FrontierMath (Feb 2025 set) | — | 5.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
| GSM8K | 92% | — |
Knowledge GPT-4.1 leads
GPT-4: 18.4 (#282), GPT-4.1: 37.1 (#160)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| GPQA Diamond | 35.7% | 66.9% |
| LMArena Expert | 1211 | 1364 |
| Humanity's Last Exam | — | 5.4% |
| SimpleQA Verified | — | 31.1% |
| MMLU-Pro | — | 81.1% |
| Vectara Hallucination Rate | — | 5.6% |
| GPQA (HELM) | — | 65.9% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multimodal Not comparable
GPT-4: —, GPT-4.1: 38.2 (#67)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| LMArena Vision | — | 1211 |
| GeoBench | — | 72% |
Multilingual GPT-4.1 leads
GPT-4: 40.6 (#215), GPT-4.1: 49.4 (#133)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| LMArena Non-English | 1246 | 1370 |
| LMArena Chinese | 1242 | 1382 |
| LMArena French | 1283 | 1382 |
| LMArena German | 1251 | 1381 |
| LMArena Japanese | 1209 | 1319 |
| LMArena Korean | 1184 | 1339 |
| LMArena Russian | 1251 | 1377 |
| LMArena Spanish | 1261 | 1376 |
Instruction Following GPT-4.1 leads
GPT-4: 65.3 (#222), GPT-4.1: 71.3 (#153)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| LMArena Instruction Following | 1241 | 1367 |
| IFEval | — | 83.8% |
Long Context GPT-4.1 leads
GPT-4: 37.7 (#212), GPT-4.1: 40.0 (#163)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| LMArena Longer Query | 1244 | 1385 |
| Fiction.LiveBench | — | 63.9% |
Writing & Preference GPT-4.1 leads
GPT-4: 34.9 (#268), GPT-4.1: 57.6 (#125)
| Benchmark | GPT-4 | GPT-4.1 |
|---|---|---|
| LMArena Text | 1263 | 1383 |
| LMArena Creative Writing | 1244 | 1363 |
| EQ-Bench Creative Writing | 752 | 1420 |
| LMArena Multi-Turn | 1257 | 1398 |
| WildBench | — | 85.4% |
Frequently asked questions
Is GPT-4 better than GPT-4.1?
GPT-4.1 is the stronger model overall, scoring 35.9 to 29.1 on the Noometry Index.
Which is cheaper, GPT-4 or GPT-4.1?
GPT-4.1 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; GPT-4 lists at $30 and $60.
Is GPT-4 or GPT-4.1 better for coding?
GPT-4.1 scores higher on coding benchmarks: 34.4 versus 31.6 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 8K.
How many benchmarks do GPT-4 and GPT-4.1 share?
27 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and GPT-4.1 has 52.