Model comparison
GPT-6 Luna vs Inkling-Small
GPT-6 Luna is the stronger model overall, scoring 53.3 to 46.5 on the Noometry Index.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-6 Luna scores higher in 7 categories and Inkling-Small in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Luna leads 76.1 to 45.1.
- The biggest single-benchmark swing is ProofBench: 64% for GPT-6 Luna and 6% for Inkling-Small.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $0.45 / $1.20 for Inkling-Small.
- GPT-6 Luna accepts more context: 1.05M tokens versus 524K.
- Inkling-Small has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Luna | Inkling-Small | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 53.3 | 46.5 |
| Released | 2026-09-22 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 524K |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $0.10 | $0.45 |
| Output $ / M tokens | $0.50 | $1.20 |
| Results tracked | 42 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Luna leads
GPT-6 Luna: 55.5 (#25), Inkling-Small: 43.6 (#85)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| LMArena WebDev | 1581 | 1409 |
| SciCode | 54.6% | 48.7% |
| LMArena Coding | 1439 | 1451 |
| DeepSWE | 66.6% | — |
| FrontierCode | 42.4% | — |
| ALE-Bench | 1,577 | — |
Agentic & Tool Use Not comparable
GPT-6 Luna: 33.3 (#54), Inkling-Small: —
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| APEX-Agents | 44.3% | — |
| GDP.pdf | 23% | — |
Reasoning GPT-6 Luna leads
GPT-6 Luna: 48.2 (#41), Inkling-Small: 38.6 (#63)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| ARC-AGI-2 | 59.3% | 40.1% |
| ARC-AGI-1 | 86.7% | 84% |
| CritPt | 19.4% | 8.3% |
| Chess Puzzles | 31% | 18% |
| LMArena Hard Prompts | 1411 | 1423 |
| Mystery Game Puzzles | 7% | 6% |
| Epoch Capabilities Index | 156.28 | 150.15 |
| NYT Connections (extended) | 68.7% | — |
| DTBench | 90.1% | — |
| LMCA | 44.5% | — |
Math GPT-6 Luna leads
GPT-6 Luna: 76.1 (#15), Inkling-Small: 45.1 (#77)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| FrontierMath (Tiers 1-3) | 78.9% | 46.3% |
| FrontierMath Tier 4 | 56.1% | 17.1% |
| OTIS Mock AIME 2024-2025 | 98.9% | 90% |
| ProofBench | 64% | 6% |
| LMArena Math | 1416 | 1459 |
Knowledge GPT-6 Luna leads
GPT-6 Luna: 57.0 (#41), Inkling-Small: 48.2 (#77)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| GPQA Diamond | 90.5% | 88.5% |
| SimpleQA Verified | 41.4% | 19.1% |
| LMArena Expert | 1444 | 1442 |
Multimodal GPT-6 Luna leads
GPT-6 Luna: 42.4 (#30), Inkling-Small: 39.1 (#62)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| LMArena Vision | 1217 | 1235 |
| Blueprint-Bench 2 | 31.2% | — |
| Furniture Assembly | 44.2% | — |
Multilingual Inkling-Small leads
GPT-6 Luna: 50.5 (#117), Inkling-Small: 51.7 (#104)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1386 | 1402 |
| LMArena Chinese | 1433 | 1465 |
| LMArena French | 1420 | 1436 |
| LMArena German | 1369 | 1405 |
| LMArena Japanese | 1369 | 1405 |
| LMArena Korean | 1360 | 1363 |
| LMArena Russian | 1394 | 1391 |
| LMArena Spanish | 1393 | 1428 |
Instruction Following Too close to call
GPT-6 Luna: 74.3 (#99), Inkling-Small: 73.8 (#114)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1409 | 1399 |
Long Context Too close to call
GPT-6 Luna: 43.0 (#111), Inkling-Small: 42.7 (#118)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1409 | 1401 |
Writing & Preference Inkling-Small leads
GPT-6 Luna: 58.3 (#119), Inkling-Small: 59.6 (#107)
| Benchmark | GPT-6 Luna | Inkling-Small |
|---|---|---|
| LMArena Text | 1391 | 1414 |
| LMArena Creative Writing | 1363 | 1331 |
| LMArena Multi-Turn | 1396 | 1418 |
| EQ-Bench Creative Writing | — | 1491 |
Frequently asked questions
Is GPT-6 Luna better than Inkling-Small?
GPT-6 Luna is the stronger model overall, scoring 53.3 to 46.5 on the Noometry Index.
Which is cheaper, GPT-6 Luna or Inkling-Small?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Inkling-Small lists at $0.45 and $1.20.
Is GPT-6 Luna or Inkling-Small better for coding?
GPT-6 Luna scores higher on coding benchmarks: 55.5 versus 43.6 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Luna does, with 1.05M tokens against 524K.
How many benchmarks do GPT-6 Luna and Inkling-Small share?
32 benchmarks have published results for both models. GPT-6 Luna has 42 scored results on Noometry and Inkling-Small has 33.