Model comparison
GPT-6 Sol vs o3
GPT-6 Sol is the stronger model overall, scoring 61.8 to 47.5 on the Noometry Index.
Last verified . 31 shared benchmarks.
Summary
- They share 31 benchmarks with published results for both. GPT-6 Sol scores higher in 8 categories and o3 in 2 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Sol leads 74.0 to 32.0.
- The biggest single-benchmark swing is ARC-AGI-2: 89.6% for GPT-6 Sol and 6.5% for o3.
- o3 is cheaper at $2 / $8 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 200K.
Side by side
| GPT-6 Sol | o3 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 61.8 | 47.5 |
| Released | 2026-09-22 | 2025-04-16 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $2 | $2 |
| Output $ / M tokens | $10 | $8 |
| Results tracked | 45 | 63 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-6 Sol: 60.1 (#11), o3: 46.8 (#64)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| LMArena Coding | 1447 | 1408 |
| ALE-Bench | 2,462 | 933.55 |
| SWE-bench Verified | — | 62.3% |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| SWE-bench Verified (bash only) | — | 58.4% |
| Aider Polyglot | — | 81.3% |
| LMArena WebDev | 1688 | — |
| SciCode | 57.6% | — |
| GSO | — | 8.8% |
| WeirdML | — | 52.4% |
| CadEval | — | 74% |
Agentic & Tool Use GPT-6 Sol leads
GPT-6 Sol: 37.2 (#36), o3: 34.5 (#44)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| APEX-Agents | 54.3% | — |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 30.8% |
| DeepResearch Bench | — | 45.2% |
| OSWorld | — | 23% |
| GDP.pdf | 26.4% | — |
| LMArena Search | — | 1144 |
| METR Time Horizons | — | 65.4% |
| Vending-Bench 2 | 14,428 | — |
Reasoning GPT-6 Sol leads
GPT-6 Sol: 74.0 (#9), o3: 32.0 (#78)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| ARC-AGI-2 | 89.6% | 6.5% |
| ARC-AGI-1 | 95.5% | 60.8% |
| CritPt | 30.9% | 1.4% |
| LMArena Hard Prompts | 1418 | 1402 |
| Mystery Game Puzzles | 56% | 29% |
| DTBench | 97.3% | 84.8% |
| LMCA | 59.1% | 39.7% |
| Epoch Capabilities Index | 162.72 | 146.86 |
| SimpleBench | — | 53.1% |
| Kagi LLM Benchmark | — | 67.6% |
| NYT Connections (extended) | 90.1% | — |
| Chess Puzzles | — | 38% |
| EnigmaEval | — | 13.1% |
| EBR-Bench | 53.3% | — |
| ForecastBench | — | 62.5 |
Math GPT-6 Sol leads
GPT-6 Sol: 87.2 (#7), o3: 50.2 (#58)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 89.8% | 33.3% |
| OTIS Mock AIME 2024-2025 | 100% | 84.4% |
| LMArena Math | 1402 | 1426 |
| FrontierMath Tier 4 | 90% | — |
| ProofBench | 83% | — |
| Omni-MATH | — | 71.4% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 18.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GPT-6 Sol leads
GPT-6 Sol: 64.8 (#15), o3: 54.6 (#52)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| GPQA Diamond | 94.3% | 81.8% |
| SimpleQA Verified | 60.7% | 49.4% |
| LMArena Expert | 1439 | 1402 |
| Humanity's Last Exam | — | 20.3% |
| MMLU-Pro | — | 85.9% |
| Confabulations | — | 14.4% |
| Vectara Hallucination Rate | 6.5% | — |
| GPQA (HELM) | — | 75.3% |
Multimodal GPT-6 Sol leads
GPT-6 Sol: 47.6 (#10), o3: 41.4 (#36)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| LMArena Vision | 1245 | 1214 |
| GeoBench | — | 74% |
| VPCT | — | 52% |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |
Multilingual o3 leads
GPT-6 Sol: 50.5 (#118), o3: 51.7 (#105)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| LMArena Non-English | 1385 | 1401 |
| LMArena Chinese | 1405 | 1437 |
| LMArena French | 1410 | 1430 |
| LMArena German | 1390 | 1420 |
| LMArena Japanese | 1385 | 1403 |
| LMArena Korean | 1341 | 1370 |
| LMArena Russian | 1401 | 1406 |
| LMArena Spanish | 1384 | 1395 |
Instruction Following GPT-6 Sol leads
GPT-6 Sol: 74.5 (#94), o3: 72.8 (#127)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| LMArena Instruction Following | 1412 | 1368 |
| IFEval | — | 86.9% |
Long Context o3 leads
GPT-6 Sol: 43.1 (#108), o3: 53.3 (#6)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| LMArena Longer Query | 1411 | 1372 |
| Fiction.LiveBench | — | 88.9% |
| CL-bench | — | 17.8% |
Writing & Preference GPT-6 Sol leads
GPT-6 Sol: 71.9 (#18), o3: 63.5 (#64)
| Benchmark | GPT-6 Sol | o3 |
|---|---|---|
| LMArena Text | 1395 | 1410 |
| LMArena Creative Writing | 1378 | 1359 |
| EQ-Bench Creative Writing | 2125 | 1676 |
| LMArena Multi-Turn | 1412 | 1405 |
| Short-Story Creative Writing | — | 83.9% |
| WildBench | — | 86.1% |
Frequently asked questions
Is GPT-6 Sol better than o3?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 47.5 on the Noometry Index.
Which is cheaper, GPT-6 Sol or o3?
o3 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is GPT-6 Sol or o3 better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 46.8 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 200K.
How many benchmarks do GPT-6 Sol and o3 share?
31 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and o3 has 63.