Model comparison
GPT-4.1 vs o1
o1 is the stronger model overall, scoring 40.9 to 35.9 on the Noometry Index. GPT-4.1 costs 7.5× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Last verified . 37 shared benchmarks.
Summary
- They share 37 benchmarks with published results for both. GPT-4.1 scores higher in 4 categories and o1 in 6 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where o1 leads 27.9 to 11.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 38.3% for GPT-4.1 and 73.3% for o1.
- GPT-4.1 is cheaper at $2 / $8 per million input/output tokens, against $15 / $60 for o1.
- GPT-4.1 accepts more context: 1.05M tokens versus 200K.
Side by side
| GPT-4.1 | o1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 35.9 | 40.9 |
| Released | 2025-04-14 | 2024-09-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 200K |
| Max output | 33K | 100K |
| Input $ / M tokens | $2 | $15 |
| Output $ / M tokens | $8 | $60 |
| Results tracked | 52 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
GPT-4.1: 34.4 (#238), o1: 46.1 (#70)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| Aider Polyglot | 52.4% | 61.7% |
| WeirdML | 39% | 47.6% |
| LMArena Coding | 1391 | 1367 |
| CadEval | 42% | 56% |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| LiveBench Coding | — | 69.7% |
| ALE-Bench | 558.1 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use GPT-4.1 leads
GPT-4.1: 34.7 (#43), o1: 24.6 (#117)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
| Cybench | — | 10% |
| METR Time Horizons | — | 51.1% |
Reasoning o1 leads
GPT-4.1: 11.7 (#339), o1: 27.9 (#111)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| SimpleBench | 27% | 41.7% |
| ARC-AGI-1 | 5.5% | 30.7% |
| Chess Puzzles | 6% | 15% |
| EnigmaEval | 2.2% | 5.7% |
| LMArena Hard Prompts | 1384 | 1371 |
| DTBench | 68.3% | 74.7% |
| LMCA | 25.6% | 22.3% |
| Epoch Capabilities Index | 136.78 | 141.91 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 52.3% | — |
| LiveBench Reasoning | — | 91.6% |
| LiveBench Data Analysis | — | 65.5% |
| ForecastBench | 61.5 | — |
| LiveBench | — | 75.7% |
Math o1 leads
GPT-4.1: 22.3 (#280), o1: 36.1 (#175)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 14.7% |
| OTIS Mock AIME 2024-2025 | 38.3% | 73.3% |
| LMArena Math | 1370 | 1388 |
| MATH Level 5 | 83% | 94.7% |
| FrontierMath (Feb 2025 set) | 5.5% | 9.3% |
| Omni-MATH | 47.1% | — |
| LiveBench Math | — | 80.3% |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge o1 leads
GPT-4.1: 37.1 (#160), o1: 41.5 (#110)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| GPQA Diamond | 66.9% | 76.8% |
| Humanity's Last Exam | 5.4% | 8% |
| SimpleQA Verified | 31.1% | 41.1% |
| LMArena Expert | 1364 | 1361 |
| MMLU-Pro | 81.1% | — |
| Confabulations | — | 11.7% |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
Multimodal GPT-4.1 leads
GPT-4.1: 38.2 (#67), o1: 34.2 (#93)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| LMArena Vision | 1211 | 1168 |
| GeoBench | 72% | 80% |
| VPCT | — | 37% |
| SpatialViz-Bench | — | 41.4% |
Multilingual Too close to call
GPT-4.1: 49.4 (#133), o1: 48.6 (#142)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| LMArena Non-English | 1370 | 1358 |
| LMArena Chinese | 1382 | 1394 |
| LMArena French | 1382 | 1344 |
| LMArena German | 1381 | 1337 |
| LMArena Japanese | 1319 | 1346 |
| LMArena Korean | 1339 | 1396 |
| LMArena Russian | 1377 | 1356 |
| LMArena Spanish | 1376 | 1345 |
Instruction Following o1 leads
GPT-4.1: 71.3 (#153), o1: 74.8 (#86)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| LMArena Instruction Following | 1367 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
| IFEval | 83.8% | — |
Long Context o1 leads
GPT-4.1: 40.0 (#163), o1: 50.3 (#9)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| Fiction.LiveBench | 63.9% | 83.3% |
| LMArena Longer Query | 1385 | 1378 |
Writing & Preference GPT-4.1 leads
GPT-4.1: 57.6 (#125), o1: 55.6 (#144)
| Benchmark | GPT-4.1 | o1 |
|---|---|---|
| LMArena Text | 1383 | 1366 |
| LMArena Creative Writing | 1363 | 1348 |
| LMArena Multi-Turn | 1398 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is GPT-4.1 better than o1?
o1 is the stronger model overall, scoring 40.9 to 35.9 on the Noometry Index. GPT-4.1 costs 7.5× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Which is cheaper, GPT-4.1 or o1?
GPT-4.1 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; o1 lists at $15 and $60.
Is GPT-4.1 or o1 better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 200K.
How many benchmarks do GPT-4.1 and o1 share?
37 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and o1 has 52.