Model comparison
GPT-6 Sol vs Grok 3
GPT-6 Sol is the stronger model overall, scoring 61.8 to 39.9 on the Noometry Index.
Last verified . 24 shared benchmarks.
Summary
- They share 24 benchmarks with published results for both. GPT-6 Sol scores higher in 7 categories and Grok 3 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Sol leads 74.0 to 13.7.
- The biggest single-benchmark swing is ARC-AGI-1: 95.5% for GPT-6 Sol and 5.5% for Grok 3.
Side by side
| GPT-6 Sol | Grok 3 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 61.8 | 39.9 |
| Released | 2026-09-22 | 2025-04-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | — |
| Max output | 128K | — |
| Input $ / M tokens | $2 | — |
| Output $ / M tokens | $10 | — |
| Results tracked | 45 | 40 |
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), Grok 3: 41.9 (#115)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| LMArena Coding | 1447 | 1432 |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| Aider Polyglot | — | 53.3% |
| LMArena WebDev | 1688 | — |
| SciCode | 57.6% | — |
| WeirdML | — | 37.2% |
| ALE-Bench | 2,462 | — |
Agentic & Tool Use GPT-6 Sol leads
GPT-6 Sol: 37.2 (#36), Grok 3: 30.5 (#76)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| APEX-Agents | 54.3% | — |
| BALROG | — | 29.5% |
| GDP.pdf | 26.4% | — |
| Vending-Bench 2 | 14,428 | — |
Reasoning GPT-6 Sol leads
GPT-6 Sol: 74.0 (#9), Grok 3: 13.7 (#333)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| ARC-AGI-2 | 89.6% | 0% |
| ARC-AGI-1 | 95.5% | 5.5% |
| LMArena Hard Prompts | 1418 | 1434 |
| Epoch Capabilities Index | 162.72 | 138.33 |
| SimpleBench | — | 36.1% |
| Kagi LLM Benchmark | — | 61.3% |
| NYT Connections (extended) | 90.1% | — |
| CritPt | 30.9% | — |
| EBR-Bench | 53.3% | — |
| Mystery Game Puzzles | 56% | — |
| DTBench | 97.3% | — |
| LMCA | 59.1% | — |
Math GPT-6 Sol leads
GPT-6 Sol: 87.2 (#7), Grok 3: 38.0 (#145)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 55.6% |
| LMArena Math | 1402 | 1391 |
| FrontierMath (Tiers 1-3) | 89.8% | — |
| FrontierMath Tier 4 | 90% | — |
| ProofBench | 83% | — |
| Omni-MATH | — | 46.4% |
| MATH Level 5 | — | 88.7% |
| FrontierMath (Feb 2025 set) | — | 3.8% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge GPT-6 Sol leads
GPT-6 Sol: 64.8 (#15), Grok 3: 46.2 (#82)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| GPQA Diamond | 94.3% | 75.8% |
| Vectara Hallucination Rate | 6.5% | 5.8% |
| LMArena Expert | 1439 | 1421 |
| SimpleQA Verified | 60.7% | — |
| MMLU-Pro | — | 78.8% |
| Confabulations | — | 14.2% |
| GPQA (HELM) | — | 65% |
Multimodal Not comparable
GPT-6 Sol: 47.6 (#10), Grok 3: —
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| LMArena Vision | 1245 | — |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |
Multilingual Grok 3 leads
GPT-6 Sol: 50.5 (#118), Grok 3: 52.3 (#87)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| LMArena Non-English | 1385 | 1410 |
| LMArena Chinese | 1405 | 1448 |
| LMArena French | 1410 | 1460 |
| LMArena German | 1390 | 1431 |
| LMArena Japanese | 1385 | 1387 |
| LMArena Korean | 1341 | 1373 |
| LMArena Russian | 1401 | 1416 |
| LMArena Spanish | 1384 | 1417 |
Instruction Following Too close to call
GPT-6 Sol: 74.5 (#94), Grok 3: 75.0 (#73)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| LMArena Instruction Following | 1412 | 1409 |
| IFEval | — | 88.4% |
Long Context GPT-6 Sol leads
GPT-6 Sol: 43.1 (#108), Grok 3: 38.7 (#192)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| LMArena Longer Query | 1411 | 1439 |
| Fiction.LiveBench | — | 58.3% |
Writing & Preference GPT-6 Sol leads
GPT-6 Sol: 71.9 (#18), Grok 3: 55.8 (#141)
| Benchmark | GPT-6 Sol | Grok 3 |
|---|---|---|
| LMArena Text | 1395 | 1426 |
| LMArena Creative Writing | 1378 | 1414 |
| EQ-Bench Creative Writing | 2125 | 1186 |
| LMArena Multi-Turn | 1412 | 1425 |
| Short-Story Creative Writing | — | 76.4% |
| WildBench | — | 84.9% |
Frequently asked questions
Is GPT-6 Sol better than Grok 3?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 39.9 on the Noometry Index.
Is GPT-6 Sol or Grok 3 better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 41.9 in the Noometry coding category.
How many benchmarks do GPT-6 Sol and Grok 3 share?
24 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and Grok 3 has 40.