# GPT-4.1 vs GPT-5.5

> GPT-5.5 is the stronger model overall, scoring 63.4 to 35.7 on the Noometry Index. GPT-4.1 costs 3.2× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-4-1-vs-gpt-5-5
- Last updated: 2026-10-10
- Shared benchmarks: 38

## Summary

- They share 38 benchmarks with published results for both. GPT-4.1 scores higher in 0 categories and GPT-5.5 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.5 leads 72.7 to 11.7.
- The biggest single-benchmark swing is ARC-AGI-1: 5.5% for GPT-4.1 and 95% for GPT-5.5.
- GPT-4.1 is cheaper at $2 / $8 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 1.05M.

## Snapshot

| | GPT-4.1 | GPT-5.5 |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 35.7 | 63.4 |
| Rank | 215 | 8 |
| Context | 1.05M | 1.05M |
| Input $/M | $2 | $5 |
| Output $/M | $8 | $30 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-4.1: 34.2 (#228)
- GPT-5.5: 58.3 (#13)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| SWE-bench Verified | 48.5% | 80.6% |
| WeirdML | 39% | 84.9% |
| LMArena Coding | 1391 | 1494 |
| ALE-Bench | 558.1 | 1,943 |
| DeepSWE | — | 67% |
| FrontierCode | — | 43% |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1512 |
| SciCode | — | 56.1% |
| GSO | — | 40.2% |
| MirrorCode | — | 10% |
| CadEval | 42% | — |

## Agentic & Tool Use

- GPT-4.1: 34.4 (#40)
- GPT-5.5: 50.7 (#6)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| Terminal-Bench | — | 84.7% |
| APEX-Agents | — | 55.1% |
| Berkeley Function Calling Leaderboard | 54% | — |
| OSWorld 2.0 | — | 13% |
| Remote Labor Index | — | 6.3% |
| τ²-bench Banking | — | 44.6% |
| DeepResearch Bench | — | 54% |
| PostTrainBench | — | 27.2% |
| ExploitBench | — | 47.4% |
| GBAEval | — | 53.2% |
| GDP.pdf | — | 26% |
| LMArena Search | — | 1242 |
| Vending-Bench 2 | — | 7,524 |

## Reasoning

- GPT-4.1: 11.7 (#334)
- GPT-5.5: 72.7 (#8)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| ARC-AGI-2 | 0.4% | 85% |
| SimpleBench | 27% | 69% |
| Kagi LLM Benchmark | 52.3% | 88.8% |
| ARC-AGI-1 | 5.5% | 95% |
| Chess Puzzles | 6% | 54% |
| LMArena Hard Prompts | 1384 | 1489 |
| DTBench | 68.3% | 96% |
| LMCA | 25.6% | 54.3% |
| Epoch Capabilities Index | 136.79 | 159.25 |
| ForecastBench | 61.5 | 60.6 |
| NYT Connections (extended) | — | 96.2% |
| CritPt | — | 27.1% |
| EnigmaEval | 2.2% | — |
| EBR-Bench | — | 34.3% |
| Mystery Game Puzzles | — | 56% |
| Surface Evolver Bench | — | 88.1% |
| Bench to the Future 3 | — | 0.14 |

## Math

- GPT-4.1: 21.4 (#269)
- GPT-5.5: 81.5 (#9)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 85.3% |
| OTIS Mock AIME 2024-2025 | 38.3% | 100% |
| LMArena Math | 1369 | 1486 |
| FrontierMath (Feb 2025 set) | 5.5% | 51.7% |
| FrontierMath Tier 4 (v1) | 0% | 35.4% |
| FrontierMath Tier 4 | — | 72.5% |
| MathArena Final-Answer Competitions | — | 94.3% |
| ProofBench | — | 50% |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath Erdős | — | 0% |

## Knowledge

- GPT-4.1: 37.0 (#152)
- GPT-5.5: 64.4 (#14)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| GPQA Diamond | 66.9% | 94% |
| SimpleQA Verified | 31.1% | 63% |
| Vectara Hallucination Rate | 5.6% | 9.3% |
| LMArena Expert | 1364 | 1508 |
| Humanity's Last Exam | 5.4% | — |
| MMLU-Pro | 81.1% | — |
| GPQA (HELM) | 65.9% | — |

## Multimodal

- GPT-4.1: 37.5 (#66)
- GPT-5.5: 47.1 (#11)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| LMArena Vision | 1210 | 1297 |
| GeoBench | 72% | — |
| Blueprint-Bench 2 | — | 36.2% |
| Furniture Assembly | — | 44.2% |
| LMArena Document | — | 1486 |

## Multilingual

- GPT-4.1: 49.6 (#128)
- GPT-5.5: 56.5 (#19)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| LMArena Non-English | 1370 | 1467 |
| LMArena Chinese | 1382 | 1535 |
| LMArena French | 1383 | 1486 |
| LMArena German | 1380 | 1478 |
| LMArena Japanese | 1320 | 1498 |
| LMArena Korean | 1339 | 1461 |
| LMArena Russian | 1377 | 1473 |
| LMArena Spanish | 1375 | 1469 |

## Instruction Following

- GPT-4.1: 71.5 (#148)
- GPT-5.5: 77.7 (#18)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| LMArena Instruction Following | 1367 | 1479 |
| IFEval | 83.8% | — |

## Long Context

- GPT-4.1: 40.0 (#158)
- GPT-5.5: 48.3 (#12)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| LMArena Longer Query | 1385 | 1484 |
| Fiction.LiveBench | 63.9% | — |
| CL-bench Life | — | 22.2% |

## Writing & Preference

- GPT-4.1: 57.8 (#120)
- GPT-5.5: 72.8 (#13)

| Benchmark | GPT-4.1 | GPT-5.5 |
|---|---|---|
| LMArena Text | 1383 | 1471 |
| LMArena Creative Writing | 1363 | 1455 |
| EQ-Bench Creative Writing | 1420 | 1844 |
| LMArena Multi-Turn | 1398 | 1477 |
| WildBench | 85.4% | — |
| EQ-Bench 4 | — | 1315 |

## FAQ

### Is GPT-4.1 better than GPT-5.5?

GPT-5.5 is the stronger model overall, scoring 63.4 to 35.7 on the Noometry Index. GPT-4.1 costs 3.2× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

### Which is cheaper, GPT-4.1 or GPT-5.5?

GPT-4.1 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; GPT-5.5 lists at $5 and $30.

### Is GPT-4.1 or GPT-5.5 better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.3 versus 34.2 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 1.05M.

### How many benchmarks do GPT-4.1 and GPT-5.5 share?

38 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and GPT-5.5 has 71.