Model comparison
GPT-4.1 vs GPT-5.1
GPT-5.1 is the stronger model overall, scoring 49.0 to 35.9 on the Noometry Index.
Last verified . 43 shared benchmarks.
Summary
- They share 43 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and GPT-5.1 in 9 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.1 leads 52.2 to 22.3.
- The biggest single-benchmark swing is ARC-AGI-1: 5.5% for GPT-4.1 and 72.8% for GPT-5.1.
- Both cost about the same: $2 input and $8 output per million tokens.
- GPT-4.1 accepts more context: 1.05M tokens versus 400K.
Side by side
| GPT-4.1 | GPT-5.1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 35.9 | 49.0 |
| Released | 2025-04-14 | 2025-11-13 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 400K |
| Max output | 33K | 128K |
| Input $ / M tokens | $2 | $1.25 |
| Output $ / M tokens | $8 | $10 |
| Results tracked | 52 | 63 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.1 leads
GPT-4.1: 34.4 (#238), GPT-5.1: 46.4 (#66)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| SWE-bench Verified | 48.5% | 68% |
| SWE-bench Verified (bash only) | 39.6% | 66% |
| WeirdML | 39% | 60.8% |
| LMArena Coding | 1391 | 1454 |
| ALE-Bench | 558.1 | 1,192 |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1395 |
| SciCode | — | 43.3% |
| GSO | — | 13.7% |
| LiveBench Coding | — | 72.5% |
| CadEval | 42% | — |
Agentic & Tool Use GPT-4.1 leads
GPT-4.1: 34.7 (#43), GPT-5.1: 32.7 (#60)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| Terminal-Bench | — | 47.6% |
| Berkeley Function Calling Leaderboard | 54% | — |
| DeepResearch Bench | — | 42.8% |
| LMArena Search | — | 1199 |
| Vending-Bench 2 | — | 1,473 |
Reasoning GPT-5.1 leads
GPT-4.1: 11.7 (#339), GPT-5.1: 39.8 (#58)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| ARC-AGI-2 | 0.4% | 17.6% |
| SimpleBench | 27% | 53.2% |
| ARC-AGI-1 | 5.5% | 72.8% |
| Chess Puzzles | 6% | 32% |
| EnigmaEval | 2.2% | 11.2% |
| LMArena Hard Prompts | 1384 | 1457 |
| DTBench | 68.3% | 90.1% |
| LMCA | 25.6% | 43.9% |
| Epoch Capabilities Index | 136.78 | 149.64 |
| ForecastBench | 61.5 | 58.1 |
| Kagi LLM Benchmark | 52.3% | — |
| CritPt | — | 4.9% |
| LiveBench Reasoning | — | 95.8% |
| Mystery Game Puzzles | — | 19% |
| LiveBench Data Analysis | — | 72.1% |
| LiveBench | — | 78.8% |
Math GPT-5.1 leads
GPT-4.1: 22.3 (#280), GPT-5.1: 52.2 (#51)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 38.3% | 88.6% |
| Omni-MATH | 47.1% | 46.4% |
| LMArena Math | 1370 | 1447 |
| FrontierMath (Feb 2025 set) | 5.5% | 31% |
| FrontierMath Tier 4 (v1) | 0% | 12.5% |
| FrontierMath (Tiers 1-3) | 6% | — |
| LiveBench Math | — | 94.5% |
| MATH Level 5 | 83% | — |
Knowledge GPT-5.1 leads
GPT-4.1: 37.1 (#160), GPT-5.1: 50.6 (#71)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| GPQA Diamond | 66.9% | 87.6% |
| Humanity's Last Exam | 5.4% | 23.7% |
| SimpleQA Verified | 31.1% | 48% |
| MMLU-Pro | 81.1% | 57.9% |
| Vectara Hallucination Rate | 5.6% | 10.9% |
| GPQA (HELM) | 65.9% | 44.2% |
| LMArena Expert | 1364 | 1470 |
Multimodal GPT-5.1 leads
GPT-4.1: 38.2 (#67), GPT-5.1: 44.8 (#19)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| LMArena Vision | 1211 | 1250 |
| GeoBench | 72% | — |
| VPCT | — | 58.7% |
| LMArena Document | — | 1403 |
Multilingual GPT-5.1 leads
GPT-4.1: 49.4 (#133), GPT-5.1: 53.8 (#56)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| LMArena Non-English | 1370 | 1431 |
| LMArena Chinese | 1382 | 1495 |
| LMArena French | 1382 | 1450 |
| LMArena German | 1381 | 1438 |
| LMArena Japanese | 1319 | 1453 |
| LMArena Korean | 1339 | 1401 |
| LMArena Russian | 1377 | 1435 |
| LMArena Spanish | 1376 | 1433 |
Instruction Following GPT-5.1 leads
GPT-4.1: 71.3 (#153), GPT-5.1: 83.9 (#1)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| IFEval | 83.8% | 93.5% |
| LMArena Instruction Following | 1367 | 1443 |
| LiveBench Instruction Following | — | 93.3% |
Long Context GPT-5.1 leads
GPT-4.1: 40.0 (#163), GPT-5.1: 47.6 (#14)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| LMArena Longer Query | 1385 | 1447 |
| Fiction.LiveBench | 63.9% | — |
| CL-bench | — | 23.7% |
| CL-bench Life | — | 17.3% |
Writing & Preference GPT-5.1 leads
GPT-4.1: 57.6 (#125), GPT-5.1: 64.5 (#55)
| Benchmark | GPT-4.1 | GPT-5.1 |
|---|---|---|
| LMArena Text | 1383 | 1443 |
| LMArena Creative Writing | 1363 | 1427 |
| WildBench | 85.4% | 86.3% |
| LMArena Multi-Turn | 1398 | 1450 |
| EQ-Bench Creative Writing | 1420 | — |
| LiveBench Language | — | 80.2% |
Frequently asked questions
Is GPT-4.1 better than GPT-5.1?
GPT-5.1 is the stronger model overall, scoring 49.0 to 35.9 on the Noometry Index.
Which is cheaper, GPT-4.1 or GPT-5.1?
GPT-5.1 is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or GPT-5.1 better for coding?
GPT-5.1 scores higher on coding benchmarks: 46.4 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 400K.
How many benchmarks do GPT-4.1 and GPT-5.1 share?
43 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and GPT-5.1 has 63.