Model comparison
GPT-5 vs Grok 3
GPT-5 is the stronger model overall, scoring 50.9 to 39.9 on the Noometry Index.
Last verified . 40 shared benchmarks.
Summary
- They share 40 benchmarks with published results for both. GPT-5 scores higher in 7 categories and Grok 3 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 38.7.
- The biggest single-benchmark swing is ARC-AGI-1: 65.7% for GPT-5 and 5.5% for Grok 3.
Side by side
| GPT-5 | Grok 3 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 50.9 | 39.9 |
| Released | 2025-08-07 | 2025-04-09 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $1.25 | — |
| Output $ / M tokens | $10 | — |
| Results tracked | 69 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
GPT-5: 50.3 (#47), Grok 3: 41.9 (#115)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| Aider Polyglot | 88% | 53.3% |
| WeirdML | 60.7% | 37.2% |
| LMArena Coding | 1436 | 1432 |
| SWE-bench Verified | 73.6% | — |
| SWE-bench Verified (bash only) | 65% | — |
| LMArena WebDev | 1418 | — |
| SciCode | 42.9% | — |
| GSO | 6.9% | — |
| ALE-Bench | 1,162 | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use GPT-5 leads
GPT-5: 33.1 (#56), Grok 3: 30.5 (#76)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| BALROG | 32.8% | 29.5% |
| Terminal-Bench | 49.6% | — |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| DeepResearch Bench | 49.6% | — |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
Reasoning GPT-5 leads
GPT-5: 38.3 (#64), Grok 3: 13.7 (#333)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| ARC-AGI-2 | 9.9% | 0% |
| SimpleBench | 56.7% | 36.1% |
| Kagi LLM Benchmark | 72.7% | 61.3% |
| ARC-AGI-1 | 65.7% | 5.5% |
| LMArena Hard Prompts | 1416 | 1434 |
| Epoch Capabilities Index | 150 | 138.33 |
| CritPt | 12.6% | — |
| Chess Puzzles | 37% | — |
| EnigmaEval | 10.5% | — |
| EBR-Bench | 12.7% | — |
| Mystery Game Puzzles | 23% | — |
| DTBench | 90.7% | — |
| LMCA | 40% | — |
| ForecastBench | 61.4 | — |
Math GPT-5 leads
GPT-5: 55.0 (#44), Grok 3: 38.0 (#145)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 55.6% |
| Omni-MATH | 64.7% | 46.4% |
| LMArena Math | 1407 | 1391 |
| MATH Level 5 | 98.1% | 88.7% |
| FrontierMath (Feb 2025 set) | 32.4% | 3.8% |
| FrontierMath Tier 4 (v1) | 12.5% | 0% |
| FrontierMath (Tiers 1-3) | 55.4% | — |
| FrontierMath Tier 4 | 22% | — |
| ProofBench | 18% | — |
Knowledge GPT-5 leads
GPT-5: 56.6 (#43), Grok 3: 46.2 (#82)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| GPQA Diamond | 86.2% | 75.8% |
| MMLU-Pro | 86.3% | 78.8% |
| Confabulations | 10.3% | 14.2% |
| Vectara Hallucination Rate | 14.7% | 5.8% |
| GPQA (HELM) | 79.2% | 65% |
| LMArena Expert | 1419 | 1421 |
| Humanity's Last Exam | 25.3% | — |
| SimpleQA Verified | 50.1% | — |
Multimodal Not comparable
GPT-5: 46.8 (#13), Grok 3: —
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| LMArena Vision | 1232 | — |
| GeoBench | 81% | — |
| VPCT | 66% | — |
Multilingual Too close to call
GPT-5: 51.4 (#110), Grok 3: 52.3 (#87)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| LMArena Non-English | 1397 | 1410 |
| LMArena Chinese | 1422 | 1448 |
| LMArena French | 1410 | 1460 |
| LMArena German | 1416 | 1431 |
| LMArena Japanese | 1409 | 1387 |
| LMArena Korean | 1360 | 1373 |
| LMArena Russian | 1406 | 1416 |
| LMArena Spanish | 1399 | 1417 |
Instruction Following Grok 3 leads
GPT-5: 73.8 (#113), Grok 3: 75.0 (#73)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| IFEval | 87.5% | 88.4% |
| LMArena Instruction Following | 1388 | 1409 |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Grok 3: 38.7 (#192)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| Fiction.LiveBench | 97.2% | 58.3% |
| LMArena Longer Query | 1399 | 1439 |
Writing & Preference GPT-5 leads
GPT-5: 63.4 (#65), Grok 3: 55.8 (#141)
| Benchmark | GPT-5 | Grok 3 |
|---|---|---|
| LMArena Text | 1406 | 1426 |
| LMArena Creative Writing | 1365 | 1414 |
| Short-Story Creative Writing | 86% | 76.4% |
| EQ-Bench Creative Writing | 1627 | 1186 |
| WildBench | 85.7% | 84.9% |
| LMArena Multi-Turn | 1426 | 1425 |
Frequently asked questions
Is GPT-5 better than Grok 3?
GPT-5 is the stronger model overall, scoring 50.9 to 39.9 on the Noometry Index.
Is GPT-5 or Grok 3 better for coding?
GPT-5 scores higher on coding benchmarks: 50.3 versus 41.9 in the Noometry coding category.
How many benchmarks do GPT-5 and Grok 3 share?
40 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Grok 3 has 40.