# GPT-5.5 vs GPT-5 Nano

> GPT-5.5 is the stronger model overall, scoring 63.4 to 33.4 on the Noometry Index. GPT-5 Nano costs 82× 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-5-5-vs-gpt-5-nano
- Last updated: 2026-10-10
- Shared benchmarks: 38

## Summary

- They share 38 benchmarks with published results for both. GPT-5.5 scores higher in 10 categories and GPT-5 Nano in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.5 leads 72.7 to 16.4.
- The biggest single-benchmark swing is ARC-AGI-2: 85% for GPT-5.5 and 2.6% for GPT-5 Nano.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 400K.

## Snapshot

| | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 63.4 | 33.4 |
| Rank | 8 | 234 |
| Context | 1.05M | 400K |
| Input $/M | $5 | $0.05 |
| Output $/M | $30 | $0.40 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-5.5: 58.3 (#13)
- GPT-5 Nano: 33.3 (#248)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| WeirdML | 84.9% | 38.1% |
| LMArena Coding | 1494 | 1351 |
| ALE-Bench | 1,943 | 718.67 |
| SWE-bench Verified | 80.6% | — |
| DeepSWE | 67% | — |
| FrontierCode | 43% | — |
| SWE-bench Verified (bash only) | — | 34.8% |
| LMArena WebDev | 1512 | — |
| SciCode | 56.1% | — |
| GSO | 40.2% | — |
| MirrorCode | 10% | — |

## Agentic & Tool Use

- GPT-5.5: 50.7 (#6)
- GPT-5 Nano: 25.6 (#99)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| Terminal-Bench | 84.7% | 21.8% |
| APEX-Agents | 55.1% | — |
| Berkeley Function Calling Leaderboard | — | 51.5% |
| 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-5.5: 72.7 (#8)
- GPT-5 Nano: 16.4 (#291)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| ARC-AGI-2 | 85% | 2.6% |
| Kagi LLM Benchmark | 88.8% | 62.2% |
| ARC-AGI-1 | 95% | 20.7% |
| Chess Puzzles | 54% | 27% |
| LMArena Hard Prompts | 1489 | 1329 |
| Mystery Game Puzzles | 56% | 9% |
| DTBench | 96% | 62.7% |
| LMCA | 54.3% | 7.9% |
| Epoch Capabilities Index | 159.25 | 139.39 |
| ForecastBench | 60.6 | 59.1 |
| SimpleBench | 69% | — |
| NYT Connections (extended) | 96.2% | — |
| CritPt | 27.1% | — |
| EBR-Bench | 34.3% | — |
| Surface Evolver Bench | 88.1% | — |
| Bench to the Future 3 | 0.14 | — |

## Math

- GPT-5.5: 81.5 (#9)
- GPT-5 Nano: 29.1 (#221)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| FrontierMath (Tiers 1-3) | 85.3% | 20% |
| FrontierMath Tier 4 | 72.5% | 2.4% |
| OTIS Mock AIME 2024-2025 | 100% | 81.1% |
| ProofBench | 50% | 12% |
| LMArena Math | 1486 | 1318 |
| FrontierMath (Feb 2025 set) | 51.7% | 8.3% |
| FrontierMath Tier 4 (v1) | 35.4% | 2.1% |
| MathArena Final-Answer Competitions | 94.3% | — |
| Omni-MATH | — | 54.6% |
| MATH Level 5 | — | 95.2% |
| FrontierMath Erdős | 0% | — |

## Knowledge

- GPT-5.5: 64.4 (#14)
- GPT-5 Nano: 35.8 (#166)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| GPQA Diamond | 94% | 69.4% |
| SimpleQA Verified | 63% | 11.7% |
| Vectara Hallucination Rate | 9.3% | 10.5% |
| LMArena Expert | 1508 | 1321 |
| MMLU-Pro | — | 77.8% |
| GPQA (HELM) | — | 67.9% |

## Multimodal

- GPT-5.5: 47.1 (#11)
- GPT-5 Nano: 30.7 (#102)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| LMArena Vision | 1297 | 1159 |
| VPCT | — | 37.2% |
| Blueprint-Bench 2 | 36.2% | — |
| Furniture Assembly | 44.2% | — |
| LMArena Document | 1486 | — |

## Multilingual

- GPT-5.5: 56.5 (#19)
- GPT-5 Nano: 45.5 (#167)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| LMArena Non-English | 1467 | 1313 |
| LMArena Chinese | 1535 | 1355 |
| LMArena German | 1478 | 1329 |
| LMArena Korean | 1461 | 1269 |
| LMArena Russian | 1473 | 1296 |
| LMArena Spanish | 1469 | 1360 |
| LMArena French | 1486 | — |
| LMArena Japanese | 1498 | — |

## Instruction Following

- GPT-5.5: 77.7 (#18)
- GPT-5 Nano: 75.2 (#72)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| LMArena Instruction Following | 1479 | 1306 |
| IFEval | — | 93.2% |

## Long Context

- GPT-5.5: 48.3 (#12)
- GPT-5 Nano: 31.3 (#275)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| LMArena Longer Query | 1484 | 1313 |
| Fiction.LiveBench | — | 44.4% |
| CL-bench Life | 22.2% | — |

## Writing & Preference

- GPT-5.5: 72.8 (#13)
- GPT-5 Nano: 39.4 (#244)

| Benchmark | GPT-5.5 | GPT-5 Nano |
|---|---|---|
| LMArena Text | 1471 | 1320 |
| LMArena Creative Writing | 1455 | 1250 |
| EQ-Bench Creative Writing | 1844 | 705 |
| LMArena Multi-Turn | 1477 | 1311 |
| WildBench | — | 80.6% |
| EQ-Bench 4 | 1315 | — |

## FAQ

### Is GPT-5.5 better than GPT-5 Nano?

GPT-5.5 is the stronger model overall, scoring 63.4 to 33.4 on the Noometry Index. GPT-5 Nano costs 82× 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-5.5 or GPT-5 Nano?

GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; GPT-5.5 lists at $5 and $30.

### Is GPT-5.5 or GPT-5 Nano better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.3 versus 33.3 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 400K.

### How many benchmarks do GPT-5.5 and GPT-5 Nano share?

38 benchmarks have published results for both models. GPT-5.5 has 71 scored results on Noometry and GPT-5 Nano has 48.