# GPT-5.6 Luna vs Inkling

> GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 44.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-5-6-luna-vs-inkling
- Last updated: 2026-10-10
- Shared benchmarks: 38

## Summary

- They share 38 benchmarks with published results for both. GPT-5.6 Luna scores higher in 8 categories and Inkling in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.5 to 31.2.
- The biggest single-benchmark swing is ProofBench: 60% for GPT-5.6 Luna and 0% for Inkling.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.6 Luna | Inkling |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 54.7 | 44.1 |
| Rank | 24 | 75 |
| Context | 1.05M | 66K |
| Input $/M | $0.20 | $1.87 |
| Output $/M | $1.20 | $4.68 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.6 Luna: 53.9 (#25)
- Inkling: 34.1 (#229)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| FrontierCode | 39.8% | 14% |
| LMArena WebDev | 1517 | 1412 |
| SciCode | 53.6% | 47% |
| WeirdML | 60.9% | 32.3% |
| LMArena Coding | 1466 | 1463 |
| ALE-Bench | 1,667 | 946 |
| DeepSWE | 67.2% | — |
| CursorBench | 35.9% | — |
| FrontierSWE | — | 4.1% |

## Agentic & Tool Use

- GPT-5.6 Luna: 35.3 (#37)
- Inkling: 29.5 (#76)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| τ²-bench Banking | — | 25% |
| BALROG | 45.6% | — |
| GDP.pdf | 22.7% | — |
| Vending-Bench 2 | 4,095 | — |

## Reasoning

- GPT-5.6 Luna: 47.5 (#38)
- Inkling: 40.6 (#48)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| ARC-AGI-2 | 59.5% | 36.5% |
| SimpleBench | 46.8% | 50% |
| ARC-AGI-1 | 88% | 79.5% |
| CritPt | 20.6% | 5.4% |
| Chess Puzzles | 40% | 21% |
| LMArena Hard Prompts | 1452 | 1452 |
| DTBench | 89.1% | 87.5% |
| LMCA | 48.5% | 37.6% |
| Epoch Capabilities Index | 156.32 | 148.6 |
| Kagi LLM Benchmark | 49.1% | — |
| NYT Connections (extended) | 69.4% | — |
| Mystery Game Puzzles | 21% | — |
| Surface Evolver Bench | 61.9% | — |

## Math

- GPT-5.6 Luna: 77.5 (#12)
- Inkling: 31.2 (#198)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.1% | 33.3% |
| FrontierMath Tier 4 | 61% | 4.9% |
| OTIS Mock AIME 2024-2025 | 98.3% | 88.9% |
| ProofBench | 60% | 0% |
| LMArena Math | 1455 | 1482 |

## Knowledge

- GPT-5.6 Luna: 58.6 (#31)
- Inkling: 55.2 (#44)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| GPQA Diamond | 91.6% | 88.3% |
| SimpleQA Verified | 41% | 40.3% |
| LMArena Expert | 1479 | 1468 |

## Multimodal

- GPT-5.6 Luna: 43.0 (#22)
- Inkling: —

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| LMArena Vision | 1259 | — |
| Blueprint-Bench 2 | 22.6% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1457 | — |

## Multilingual

- GPT-5.6 Luna: 53.0 (#73)
- Inkling: 54.2 (#49)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| LMArena Non-English | 1418 | 1434 |
| LMArena Chinese | 1469 | 1495 |
| LMArena French | 1457 | 1453 |
| LMArena German | 1459 | 1456 |
| LMArena Japanese | 1416 | 1428 |
| LMArena Korean | 1414 | 1401 |
| LMArena Russian | 1428 | 1430 |
| LMArena Spanish | 1445 | 1442 |

## Instruction Following

- GPT-5.6 Luna: 76.0 (#48)
- Inkling: 75.3 (#65)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| LMArena Instruction Following | 1440 | 1427 |

## Long Context

- GPT-5.6 Luna: 43.9 (#78)
- Inkling: 43.9 (#79)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| LMArena Longer Query | 1437 | 1436 |

## Writing & Preference

- GPT-5.6 Luna: 68.3 (#27)
- Inkling: 65.5 (#48)

| Benchmark | GPT-5.6 Luna | Inkling |
|---|---|---|
| LMArena Text | 1432 | 1443 |
| LMArena Creative Writing | 1397 | 1390 |
| EQ-Bench Creative Writing | 1829 | 1611 |
| EQ-Bench 4 | 1156 | 1226 |
| LMArena Multi-Turn | 1433 | 1436 |

## FAQ

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

GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 44.1 on the Noometry Index.

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

GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; Inkling lists at $1.87 and $4.68.

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

GPT-5.6 Luna scores higher on coding benchmarks: 53.9 versus 34.1 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Luna does, with 1.05M tokens against 66K.

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

38 benchmarks have published results for both models. GPT-5.6 Luna has 51 scored results on Noometry and Inkling has 41.