Model comparison
GPT-6 Luna vs GPT-6 Sol
GPT-6 Sol is the stronger model overall, scoring 61.8 to 53.3 on the Noometry Index. GPT-6 Luna costs 20× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Last verified . 41 shared benchmarks.
Summary
- They share 41 benchmarks with published results for both. GPT-6 Luna scores higher in 1 category and GPT-6 Sol in 9 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Sol leads 74.0 to 48.2.
- The biggest single-benchmark swing is Mystery Game Puzzles: 7% for GPT-6 Luna and 56% for GPT-6 Sol.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
Side by side
| GPT-6 Luna | GPT-6 Sol | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 53.3 | 61.8 |
| Released | 2026-09-22 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $0.10 | $2 |
| Output $ / M tokens | $0.50 | $10 |
| Results tracked | 42 | 45 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-6 Luna: 55.5 (#25), GPT-6 Sol: 60.1 (#11)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| DeepSWE | 66.6% | 68.8% |
| FrontierCode | 42.4% | 49.3% |
| LMArena WebDev | 1581 | 1688 |
| SciCode | 54.6% | 57.6% |
| LMArena Coding | 1439 | 1447 |
| ALE-Bench | 1,577 | 2,462 |
Agentic & Tool Use GPT-6 Sol leads
GPT-6 Luna: 33.3 (#54), GPT-6 Sol: 37.2 (#36)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| APEX-Agents | 44.3% | 54.3% |
| GDP.pdf | 23% | 26.4% |
| Vending-Bench 2 | — | 14,428 |
Reasoning GPT-6 Sol leads
GPT-6 Luna: 48.2 (#41), GPT-6 Sol: 74.0 (#9)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| ARC-AGI-2 | 59.3% | 89.6% |
| NYT Connections (extended) | 68.7% | 90.1% |
| ARC-AGI-1 | 86.7% | 95.5% |
| CritPt | 19.4% | 30.9% |
| LMArena Hard Prompts | 1411 | 1418 |
| Mystery Game Puzzles | 7% | 56% |
| DTBench | 90.1% | 97.3% |
| LMCA | 44.5% | 59.1% |
| Epoch Capabilities Index | 156.28 | 162.72 |
| Chess Puzzles | 31% | — |
| EBR-Bench | — | 53.3% |
Math GPT-6 Sol leads
GPT-6 Luna: 76.1 (#15), GPT-6 Sol: 87.2 (#7)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| FrontierMath (Tiers 1-3) | 78.9% | 89.8% |
| FrontierMath Tier 4 | 56.1% | 90% |
| OTIS Mock AIME 2024-2025 | 98.9% | 100% |
| ProofBench | 64% | 83% |
| LMArena Math | 1416 | 1402 |
Knowledge GPT-6 Sol leads
GPT-6 Luna: 57.0 (#41), GPT-6 Sol: 64.8 (#15)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| GPQA Diamond | 90.5% | 94.3% |
| SimpleQA Verified | 41.4% | 60.7% |
| LMArena Expert | 1444 | 1439 |
| Vectara Hallucination Rate | — | 6.5% |
Multimodal GPT-6 Sol leads
GPT-6 Luna: 42.4 (#30), GPT-6 Sol: 47.6 (#10)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| LMArena Vision | 1217 | 1245 |
| Blueprint-Bench 2 | 31.2% | 36.9% |
| Furniture Assembly | 44.2% | 58.3% |
Multilingual Too close to call
GPT-6 Luna: 50.5 (#117), GPT-6 Sol: 50.5 (#118)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| LMArena Non-English | 1386 | 1385 |
| LMArena Chinese | 1433 | 1405 |
| LMArena French | 1420 | 1410 |
| LMArena German | 1369 | 1390 |
| LMArena Japanese | 1369 | 1385 |
| LMArena Korean | 1360 | 1341 |
| LMArena Russian | 1394 | 1401 |
| LMArena Spanish | 1393 | 1384 |
Instruction Following Too close to call
GPT-6 Luna: 74.3 (#99), GPT-6 Sol: 74.5 (#94)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| LMArena Instruction Following | 1409 | 1412 |
Long Context Too close to call
GPT-6 Luna: 43.0 (#111), GPT-6 Sol: 43.1 (#108)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| LMArena Longer Query | 1409 | 1411 |
Writing & Preference GPT-6 Sol leads
GPT-6 Luna: 58.3 (#119), GPT-6 Sol: 71.9 (#18)
| Benchmark | GPT-6 Luna | GPT-6 Sol |
|---|---|---|
| LMArena Text | 1391 | 1395 |
| LMArena Creative Writing | 1363 | 1378 |
| LMArena Multi-Turn | 1396 | 1412 |
| EQ-Bench Creative Writing | — | 2125 |
Frequently asked questions
Is GPT-6 Luna better than GPT-6 Sol?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 53.3 on the Noometry Index. GPT-6 Luna costs 20× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Luna or GPT-6 Sol?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is GPT-6 Luna or GPT-6 Sol better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
Both accept 1.05M tokens.
How many benchmarks do GPT-6 Luna and GPT-6 Sol share?
41 benchmarks have published results for both models. GPT-6 Luna has 42 scored results on Noometry and GPT-6 Sol has 45.