Model comparison
GPT-5.6 Sol vs GPT-6 Luna
GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 53.3 on the Noometry Index. GPT-6 Luna costs 40× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Last verified . 42 shared benchmarks.
Summary
- They share 42 benchmarks with published results for both. GPT-5.6 Sol scores higher in 10 categories and GPT-6 Luna in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Sol leads 74.8 to 48.2.
- The biggest single-benchmark swing is Mystery Game Puzzles: 58% for GPT-5.6 Sol and 7% for GPT-6 Luna.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
Side by side
| GPT-5.6 Sol | GPT-6 Luna | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 65.0 | 53.3 |
| Released | 2026-07-09 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $4 | $0.10 |
| Output $ / M tokens | $20 | $0.50 |
| Results tracked | 65 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Sol leads
GPT-5.6 Sol: 65.1 (#7), GPT-6 Luna: 55.5 (#25)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| DeepSWE | 72.7% | 66.6% |
| FrontierCode | 47.5% | 42.4% |
| LMArena WebDev | 1618 | 1581 |
| SciCode | 57.1% | 54.6% |
| LMArena Coding | 1498 | 1439 |
| ALE-Bench | 2,177 | 1,577 |
| CursorBench | 41.7% | — |
| FrontierSWE | 32.2% | — |
| GSO | 76.5% | — |
| WeirdML | 89.4% | — |
| MirrorCode | 20% | — |
Agentic & Tool Use GPT-5.6 Sol leads
GPT-5.6 Sol: 50.3 (#7), GPT-6 Luna: 33.3 (#54)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| APEX-Agents | 51.4% | 44.3% |
| GDP.pdf | 30.7% | 23% |
| OSWorld 2.0 | 27.3% | — |
| τ²-bench Banking | 46.9% | — |
| PostTrainBench | 36.2% | — |
| BALROG | 60% | — |
| GBAEval | 52.6% | — |
| LMArena Search | 1257 | — |
| Vending-Bench 2 | 9,619 | — |
Reasoning GPT-5.6 Sol leads
GPT-5.6 Sol: 74.8 (#8), GPT-6 Luna: 48.2 (#41)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| ARC-AGI-2 | 92.5% | 59.3% |
| NYT Connections (extended) | 93.8% | 68.7% |
| ARC-AGI-1 | 97.5% | 86.7% |
| CritPt | 32.3% | 19.4% |
| Chess Puzzles | 64% | 31% |
| LMArena Hard Prompts | 1484 | 1411 |
| Mystery Game Puzzles | 58% | 7% |
| DTBench | 96% | 90.1% |
| LMCA | 59.2% | 44.5% |
| Epoch Capabilities Index | 161.66 | 156.28 |
| SimpleBench | 71.7% | — |
| Kagi LLM Benchmark | 67% | — |
| EnigmaEval | 37.1% | — |
| EBR-Bench | 44.8% | — |
| Surface Evolver Bench | 93.1% | — |
| Bench to the Future 3 | 0.14 | — |
Math GPT-5.6 Sol leads
GPT-5.6 Sol: 85.6 (#9), GPT-6 Luna: 76.1 (#15)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| FrontierMath (Tiers 1-3) | 89.1% | 78.9% |
| FrontierMath Tier 4 | 82.9% | 56.1% |
| OTIS Mock AIME 2024-2025 | 100% | 98.9% |
| ProofBench | 83% | 64% |
| LMArena Math | 1474 | 1416 |
| FrontierMath Erdős | 0% | — |
Knowledge GPT-5.6 Sol leads
GPT-5.6 Sol: 64.3 (#18), GPT-6 Luna: 57.0 (#41)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| GPQA Diamond | 93.5% | 90.5% |
| SimpleQA Verified | 69.7% | 41.4% |
| LMArena Expert | 1516 | 1444 |
| Vectara Hallucination Rate | 12.4% | — |
Multimodal GPT-5.6 Sol leads
GPT-5.6 Sol: 48.6 (#9), GPT-6 Luna: 42.4 (#30)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| LMArena Vision | 1281 | 1217 |
| Blueprint-Bench 2 | 33.6% | 31.2% |
| Furniture Assembly | 56.7% | 44.2% |
| LMArena Document | 1483 | — |
Multilingual GPT-5.6 Sol leads
GPT-5.6 Sol: 55.3 (#32), GPT-6 Luna: 50.5 (#117)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| LMArena Non-English | 1452 | 1386 |
| LMArena Chinese | 1527 | 1433 |
| LMArena French | 1477 | 1420 |
| LMArena German | 1476 | 1369 |
| LMArena Japanese | 1471 | 1369 |
| LMArena Korean | 1442 | 1360 |
| LMArena Russian | 1468 | 1394 |
| LMArena Spanish | 1441 | 1393 |
Instruction Following GPT-5.6 Sol leads
GPT-5.6 Sol: 77.7 (#16), GPT-6 Luna: 74.3 (#99)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| LMArena Instruction Following | 1482 | 1409 |
Long Context GPT-5.6 Sol leads
GPT-5.6 Sol: 45.4 (#42), GPT-6 Luna: 43.0 (#111)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| LMArena Longer Query | 1480 | 1409 |
Writing & Preference GPT-5.6 Sol leads
GPT-5.6 Sol: 73.3 (#12), GPT-6 Luna: 58.3 (#119)
| Benchmark | GPT-5.6 Sol | GPT-6 Luna |
|---|---|---|
| LMArena Text | 1457 | 1391 |
| LMArena Creative Writing | 1448 | 1363 |
| LMArena Multi-Turn | 1460 | 1396 |
| EQ-Bench Creative Writing | 1972 | — |
| EQ-Bench 4 | 1250 | — |
Frequently asked questions
Is GPT-5.6 Sol better than GPT-6 Luna?
GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 53.3 on the Noometry Index. GPT-6 Luna costs 40× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Sol or GPT-6 Luna?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; GPT-5.6 Sol lists at $4 and $20.
Is GPT-5.6 Sol or GPT-6 Luna better for coding?
GPT-5.6 Sol scores higher on coding benchmarks: 65.1 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
Both accept 1.05M tokens.
How many benchmarks do GPT-5.6 Sol and GPT-6 Luna share?
42 benchmarks have published results for both models. GPT-5.6 Sol has 65 scored results on Noometry and GPT-6 Luna has 42.