Model comparison
GPT-4 vs GPT-6 Sol
GPT-6 Sol is the stronger model overall, scoring 61.8 to 29.1 on the Noometry Index.
Last verified . 24 shared benchmarks.
Summary
- They share 24 benchmarks with published results for both. GPT-4 scores higher in 0 categories and GPT-6 Sol in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Sol leads 87.2 to 10.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.1% for GPT-4 and 100% for GPT-6 Sol.
- GPT-6 Sol is cheaper at $2 / $10 per million input/output tokens, against $30 / $60 for GPT-4.
- GPT-6 Sol accepts more context: 1.05M tokens versus 8K.
Side by side
| GPT-4 | GPT-6 Sol | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 29.1 | 61.8 |
| Released | 2023-03-14 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 8K | 1.05M |
| Max output | 8K | 128K |
| Input $ / M tokens | $30 | $2 |
| Output $ / M tokens | $60 | $10 |
| Results tracked | 38 | 45 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-4: 31.6 (#283), GPT-6 Sol: 60.1 (#11)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Coding | 1254 | 1447 |
| DeepSWE | — | 68.8% |
| FrontierCode | — | 49.3% |
| LMArena WebDev | — | 1688 |
| SciCode | — | 57.6% |
| WeirdML | 12.4% | — |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| ALE-Bench | — | 2,462 |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, GPT-6 Sol: 37.2 (#36)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| APEX-Agents | — | 54.3% |
| GDP.pdf | — | 26.4% |
| METR Time Horizons | 36.1% | — |
| Vending-Bench 2 | — | 14,428 |
Reasoning GPT-6 Sol leads
GPT-4: 17.8 (#289), GPT-6 Sol: 74.0 (#9)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1418 |
| Mystery Game Puzzles | 12% | 56% |
| DTBench | 62.7% | 97.3% |
| LMCA | 17.1% | 59.1% |
| Epoch Capabilities Index | 125.89 | 162.72 |
| ARC-AGI-2 | — | 89.6% |
| NYT Connections (extended) | — | 90.1% |
| ARC-AGI-1 | — | 95.5% |
| CritPt | — | 30.9% |
| Chess Puzzles | 4% | — |
| EBR-Bench | — | 53.3% |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math GPT-6 Sol leads
GPT-4: 10.8 (#309), GPT-6 Sol: 87.2 (#7)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 100% |
| LMArena Math | 1269 | 1402 |
| FrontierMath (Tiers 1-3) | — | 89.8% |
| FrontierMath Tier 4 | — | 90% |
| ProofBench | — | 83% |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge GPT-6 Sol leads
GPT-4: 18.4 (#282), GPT-6 Sol: 64.8 (#15)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| GPQA Diamond | 35.7% | 94.3% |
| LMArena Expert | 1211 | 1439 |
| SimpleQA Verified | — | 60.7% |
| Vectara Hallucination Rate | — | 6.5% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multimodal Not comparable
GPT-4: —, GPT-6 Sol: 47.6 (#10)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Vision | — | 1245 |
| Blueprint-Bench 2 | — | 36.9% |
| Furniture Assembly | — | 58.3% |
Multilingual GPT-6 Sol leads
GPT-4: 40.6 (#215), GPT-6 Sol: 50.5 (#118)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Non-English | 1246 | 1385 |
| LMArena Chinese | 1242 | 1405 |
| LMArena French | 1283 | 1410 |
| LMArena German | 1251 | 1390 |
| LMArena Japanese | 1209 | 1385 |
| LMArena Korean | 1184 | 1341 |
| LMArena Russian | 1251 | 1401 |
| LMArena Spanish | 1261 | 1384 |
Instruction Following GPT-6 Sol leads
GPT-4: 65.3 (#222), GPT-6 Sol: 74.5 (#94)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Instruction Following | 1241 | 1412 |
Long Context GPT-6 Sol leads
GPT-4: 37.7 (#212), GPT-6 Sol: 43.1 (#108)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Longer Query | 1244 | 1411 |
Writing & Preference GPT-6 Sol leads
GPT-4: 34.9 (#268), GPT-6 Sol: 71.9 (#18)
| Benchmark | GPT-4 | GPT-6 Sol |
|---|---|---|
| LMArena Text | 1263 | 1395 |
| LMArena Creative Writing | 1244 | 1378 |
| EQ-Bench Creative Writing | 752 | 2125 |
| LMArena Multi-Turn | 1257 | 1412 |
Frequently asked questions
Is GPT-4 better than GPT-6 Sol?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 29.1 on the Noometry Index.
Which is cheaper, GPT-4 or GPT-6 Sol?
GPT-6 Sol is cheaper. It lists at $2 per million input tokens and $10 per million output tokens; GPT-4 lists at $30 and $60.
Is GPT-4 or GPT-6 Sol better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 31.6 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 8K.
How many benchmarks do GPT-4 and GPT-6 Sol share?
24 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and GPT-6 Sol has 45.