# GPT-4.1 vs o3

> o3 is the stronger model overall, scoring 47.5 to 35.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-1-vs-o3
- Last updated: 2026-10-10
- Shared benchmarks: 51

## Summary

- They share 51 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and o3 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where o3 leads 50.2 to 22.3.
- The biggest single-benchmark swing is ARC-AGI-1: 5.5% for GPT-4.1 and 60.8% for o3.
- Both cost about the same: $2 input and $8 output per million tokens.
- GPT-4.1 accepts more context: 1.05M tokens versus 200K.

## Snapshot

| | GPT-4.1 | o3 |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 35.9 | 47.5 |
| Rank | 219 | 61 |
| Context | 1.05M | 200K |
| Input $/M | $2 | $2 |
| Output $/M | $8 | $8 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-4.1: 34.4 (#238)
- o3: 46.8 (#64)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| SWE-bench Verified | 48.5% | 62.3% |
| SWE-bench Verified (bash only) | 39.6% | 58.4% |
| Aider Polyglot | 52.4% | 81.3% |
| WeirdML | 39% | 52.4% |
| LMArena Coding | 1391 | 1408 |
| CadEval | 42% | 74% |
| ALE-Bench | 558.1 | 933.55 |
| GSO | — | 8.8% |

## Agentic & Tool Use

- GPT-4.1: 34.7 (#43)
- o3: 34.5 (#44)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | 63% |
| GDPval | — | 30.8% |
| DeepResearch Bench | — | 45.2% |
| OSWorld | — | 23% |
| LMArena Search | — | 1144 |
| METR Time Horizons | — | 65.4% |

## Reasoning

- GPT-4.1: 11.7 (#339)
- o3: 32.0 (#78)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| ARC-AGI-2 | 0.4% | 6.5% |
| SimpleBench | 27% | 53.1% |
| Kagi LLM Benchmark | 52.3% | 67.6% |
| ARC-AGI-1 | 5.5% | 60.8% |
| Chess Puzzles | 6% | 38% |
| EnigmaEval | 2.2% | 13.1% |
| LMArena Hard Prompts | 1384 | 1402 |
| DTBench | 68.3% | 84.8% |
| LMCA | 25.6% | 39.7% |
| Epoch Capabilities Index | 136.78 | 146.86 |
| ForecastBench | 61.5 | 62.5 |
| CritPt | — | 1.4% |
| Mystery Game Puzzles | — | 29% |

## Math

- GPT-4.1: 22.3 (#280)
- o3: 50.2 (#58)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 33.3% |
| OTIS Mock AIME 2024-2025 | 38.3% | 84.4% |
| Omni-MATH | 47.1% | 71.4% |
| LMArena Math | 1370 | 1426 |
| MATH Level 5 | 83% | 97.8% |
| FrontierMath (Feb 2025 set) | 5.5% | 18.7% |
| FrontierMath Tier 4 (v1) | 0% | 2.1% |

## Knowledge

- GPT-4.1: 37.1 (#160)
- o3: 54.6 (#52)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| GPQA Diamond | 66.9% | 81.8% |
| Humanity's Last Exam | 5.4% | 20.3% |
| SimpleQA Verified | 31.1% | 49.4% |
| MMLU-Pro | 81.1% | 85.9% |
| GPQA (HELM) | 65.9% | 75.3% |
| LMArena Expert | 1364 | 1402 |
| Confabulations | — | 14.4% |
| Vectara Hallucination Rate | 5.6% | — |

## Multimodal

- GPT-4.1: 38.2 (#67)
- o3: 41.4 (#36)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| LMArena Vision | 1211 | 1214 |
| GeoBench | 72% | 74% |
| VPCT | — | 52% |

## Multilingual

- GPT-4.1: 49.4 (#133)
- o3: 51.7 (#105)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| LMArena Non-English | 1370 | 1401 |
| LMArena Chinese | 1382 | 1437 |
| LMArena French | 1382 | 1430 |
| LMArena German | 1381 | 1420 |
| LMArena Japanese | 1319 | 1403 |
| LMArena Korean | 1339 | 1370 |
| LMArena Russian | 1377 | 1406 |
| LMArena Spanish | 1376 | 1395 |

## Instruction Following

- GPT-4.1: 71.3 (#153)
- o3: 72.8 (#127)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| IFEval | 83.8% | 86.9% |
| LMArena Instruction Following | 1367 | 1368 |

## Long Context

- GPT-4.1: 40.0 (#163)
- o3: 53.3 (#6)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| Fiction.LiveBench | 63.9% | 88.9% |
| LMArena Longer Query | 1385 | 1372 |
| CL-bench | — | 17.8% |

## Writing & Preference

- GPT-4.1: 57.6 (#125)
- o3: 63.5 (#64)

| Benchmark | GPT-4.1 | o3 |
|---|---|---|
| LMArena Text | 1383 | 1410 |
| LMArena Creative Writing | 1363 | 1359 |
| EQ-Bench Creative Writing | 1420 | 1676 |
| WildBench | 85.4% | 86.1% |
| LMArena Multi-Turn | 1398 | 1405 |
| Short-Story Creative Writing | — | 83.9% |

## FAQ

### Is GPT-4.1 better than o3?

o3 is the stronger model overall, scoring 47.5 to 35.9 on the Noometry Index.

### Which is cheaper, GPT-4.1 or o3?

o3 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; GPT-4.1 lists at $2 and $8.

### Is GPT-4.1 or o3 better for coding?

o3 scores higher on coding benchmarks: 46.8 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 o3 share?

51 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and o3 has 63.