# GPT-5.6 Luna vs o1

> GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 40.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-5-6-luna-vs-o1
- Last updated: 2026-10-10
- Shared benchmarks: 29

## Summary

- They share 29 benchmarks with published results for both. GPT-5.6 Luna scores higher in 9 categories and o1 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.5 to 34.9.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 82.1% for GPT-5.6 Luna and 14.7% for o1.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $15 / $60 for o1.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 200K.

## Snapshot

| | GPT-5.6 Luna | o1 |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 54.7 | 40.6 |
| Rank | 24 | 140 |
| Context | 1.05M | 200K |
| Input $/M | $0.20 | $15 |
| Output $/M | $1.20 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-5.6 Luna: 53.9 (#25)
- o1: 45.7 (#64)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| WeirdML | 60.9% | 47.6% |
| LMArena Coding | 1466 | 1367 |
| DeepSWE | 67.2% | — |
| FrontierCode | 39.8% | — |
| Aider Polyglot | — | 61.7% |
| CursorBench | 35.9% | — |
| LMArena WebDev | 1517 | — |
| SciCode | 53.6% | — |
| LiveBench Coding | — | 69.7% |
| CadEval | — | 56% |
| ALE-Bench | 1,667 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |

## Agentic & Tool Use

- GPT-5.6 Luna: 35.3 (#37)
- o1: 24.3 (#111)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| Cybench | — | 10% |
| BALROG | 45.6% | — |
| GDP.pdf | 22.7% | — |
| METR Time Horizons | — | 51.1% |
| Vending-Bench 2 | 4,095 | — |

## Reasoning

- GPT-5.6 Luna: 47.5 (#38)
- o1: 27.5 (#103)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| SimpleBench | 46.8% | 41.7% |
| ARC-AGI-1 | 88% | 30.7% |
| Chess Puzzles | 40% | 15% |
| LMArena Hard Prompts | 1452 | 1372 |
| DTBench | 89.1% | 74.7% |
| LMCA | 48.5% | 22.3% |
| Epoch Capabilities Index | 156.32 | 141.92 |
| ARC-AGI-2 | 59.5% | — |
| Kagi LLM Benchmark | 49.1% | — |
| NYT Connections (extended) | 69.4% | — |
| CritPt | 20.6% | — |
| EnigmaEval | — | 5.7% |
| LiveBench Reasoning | — | 91.6% |
| Mystery Game Puzzles | 21% | — |
| LiveBench Data Analysis | — | 65.5% |
| Surface Evolver Bench | 61.9% | — |
| LiveBench | — | 75.7% |

## Math

- GPT-5.6 Luna: 77.5 (#12)
- o1: 34.9 (#150)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.1% | 14.7% |
| OTIS Mock AIME 2024-2025 | 98.3% | 73.3% |
| LMArena Math | 1455 | 1388 |
| FrontierMath Tier 4 | 61% | — |
| ProofBench | 60% | — |
| LiveBench Math | — | 80.3% |
| MATH Level 5 | — | 94.7% |
| FrontierMath (Feb 2025 set) | — | 9.3% |

## Knowledge

- GPT-5.6 Luna: 58.6 (#31)
- o1: 41.5 (#101)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| GPQA Diamond | 91.6% | 76.8% |
| SimpleQA Verified | 41% | 41.1% |
| LMArena Expert | 1479 | 1362 |
| Humanity's Last Exam | — | 8% |
| Confabulations | — | 11.7% |

## Multimodal

- GPT-5.6 Luna: 43.0 (#22)
- o1: 33.6 (#87)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| LMArena Vision | 1259 | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| Blueprint-Bench 2 | 22.6% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1457 | — |
| SpatialViz-Bench | — | 41.4% |

## Multilingual

- GPT-5.6 Luna: 53.0 (#73)
- o1: 48.8 (#137)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| LMArena Non-English | 1418 | 1359 |
| LMArena Chinese | 1469 | 1394 |
| LMArena French | 1457 | 1343 |
| LMArena German | 1459 | 1337 |
| LMArena Japanese | 1416 | 1346 |
| LMArena Korean | 1414 | 1396 |
| LMArena Russian | 1428 | 1356 |
| LMArena Spanish | 1445 | 1345 |

## Instruction Following

- GPT-5.6 Luna: 76.0 (#48)
- o1: 75.0 (#82)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| LMArena Instruction Following | 1440 | 1368 |
| LiveBench Instruction Following | — | 81.5% |

## Long Context

- GPT-5.6 Luna: 43.9 (#78)
- o1: 50.3 (#9)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| LMArena Longer Query | 1437 | 1378 |
| Fiction.LiveBench | — | 83.3% |

## Writing & Preference

- GPT-5.6 Luna: 68.3 (#27)
- o1: 55.8 (#139)

| Benchmark | GPT-5.6 Luna | o1 |
|---|---|---|
| LMArena Text | 1432 | 1366 |
| LMArena Creative Writing | 1397 | 1347 |
| LMArena Multi-Turn | 1433 | 1368 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 1829 | — |
| EQ-Bench 4 | 1156 | — |
| LiveBench Language | — | 65.4% |

## FAQ

### Is GPT-5.6 Luna better than o1?

GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 40.6 on the Noometry Index.

### Which is cheaper, GPT-5.6 Luna or o1?

GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; o1 lists at $15 and $60.

### Is GPT-5.6 Luna or o1 better for coding?

GPT-5.6 Luna scores higher on coding benchmarks: 53.9 versus 45.7 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Luna does, with 1.05M tokens against 200K.

### How many benchmarks do GPT-5.6 Luna and o1 share?

29 benchmarks have published results for both models. GPT-5.6 Luna has 51 scored results on Noometry and o1 has 52.