Model comparison
GPT-4o vs Grok 4.6
Grok 4.6 is the stronger model overall, scoring 56.9 to 28.6 on the Noometry Index.
Last verified . 31 shared benchmarks.
Summary
- They share 31 benchmarks with published results for both. GPT-4o scores higher in 0 categories and Grok 4.6 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4.6 leads 67.0 to 10.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.4% for GPT-4o and 99.2% for Grok 4.6.
- Grok 4.6 is cheaper at $2 / $6 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- Grok 4.6 accepts more context: 500K tokens versus 128K.
Side by side
| GPT-4o | Grok 4.6 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 28.6 | 56.9 |
| Released | 2024-05-13 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 128K | 500K |
| Max output | 16K | 500K |
| Input $ / M tokens | $2.50 | $2 |
| Output $ / M tokens | $10 | $6 |
| Results tracked | 72 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.6 leads
GPT-4o: 24.8 (#328), Grok 4.6: 58.5 (#16)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| WeirdML | 25.1% | 67.3% |
| LMArena Coding | 1297 | 1465 |
| SWE-bench Verified | 31% | — |
| DeepSWE | — | 67.5% |
| FrontierCode | — | 48% |
| SWE-bench Verified (bash only) | 21.6% | — |
| Aider Polyglot | 45.3% | — |
| CursorBench | — | 41.4% |
| LMArena WebDev | — | 1617 |
| FrontierSWE | — | 25.3% |
| SciCode | — | 56.5% |
| GSO | 0% | — |
| BigCodeBench Instruct | 51.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 61.1% | — |
| CadEval | 26% | — |
| ALE-Bench | — | 1,508 |
| HumanEval+ | 87.2% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Grok 4.6 leads
GPT-4o: 21.0 (#141), Grok 4.6: 39.4 (#27)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| APEX-Agents | — | 65.3% |
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| GDP.pdf | — | 17.2% |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
| Vending-Bench 2 | — | 9,047 |
Reasoning Grok 4.6 leads
GPT-4o: 9.4 (#343), Grok 4.6: 61.4 (#20)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| ARC-AGI-2 | 0% | 67.1% |
| SimpleBench | 17.8% | 75.9% |
| ARC-AGI-1 | 4.5% | 87.5% |
| CritPt | 0% | 19.7% |
| Chess Puzzles | 13% | 40% |
| LMArena Hard Prompts | 1281 | 1447 |
| DTBench | 64.5% | 97.3% |
| LMCA | 16.6% | 48.5% |
| Epoch Capabilities Index | 128.97 | 156.44 |
| NYT Connections (extended) | — | 80% |
| EnigmaEval | 0.8% | — |
| EBR-Bench | — | 30.5% |
| LiveBench Reasoning | 55.8% | — |
| Mystery Game Puzzles | — | 34% |
| LiveBench Data Analysis | 60.9% | — |
| ForecastBench | 57.7 | — |
| LiveBench | 55.3% | — |
Math Grok 4.6 leads
GPT-4o: 10.6 (#312), Grok 4.6: 67.0 (#24)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.4% | 66% |
| OTIS Mock AIME 2024-2025 | 6.4% | 99.2% |
| LMArena Math | 1285 | 1423 |
| FrontierMath Tier 4 | — | 31.7% |
| ProofBench | — | 51% |
| Omni-MATH | 29.3% | — |
| LiveBench Math | 49.5% | — |
| MATH Level 5 | 53.3% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Grok 4.6 leads
GPT-4o: 28.8 (#242), Grok 4.6: 63.3 (#20)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 49.2% | 94% |
| SimpleQA Verified | 26% | 49.3% |
| LMArena Expert | 1250 | 1467 |
| Humanity's Last Exam | 2.7% | — |
| MMLU-Pro | 71.3% | — |
| Confabulations | 15.3% | — |
| Vectara Hallucination Rate | 9.6% | — |
| GPQA (HELM) | 52% | — |
| MMLU | 88.1% | — |
Multimodal Grok 4.6 leads
GPT-4o: 34.5 (#91), Grok 4.6: 43.6 (#23)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1137 | 1263 |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| Blueprint-Bench 2 | — | 33.2% |
| Furniture Assembly | — | 40% |
| LMArena Document | — | 1452 |
| ScienceQA | 88.5% | — |
Multilingual Grok 4.6 leads
GPT-4o: 43.2 (#186), Grok 4.6: 53.0 (#74)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1283 | 1420 |
| LMArena Chinese | 1277 | 1480 |
| LMArena French | 1304 | 1461 |
| LMArena German | 1282 | 1431 |
| LMArena Japanese | 1257 | 1376 |
| LMArena Korean | 1234 | 1397 |
| LMArena Russian | 1286 | 1422 |
| LMArena Spanish | 1292 | 1404 |
Instruction Following Grok 4.6 leads
GPT-4o: 66.6 (#207), Grok 4.6: 75.4 (#63)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1278 | 1431 |
| LiveBench Instruction Following | 68.6% | — |
| IFEval | 81.7% | — |
Long Context Grok 4.6 leads
GPT-4o: 39.4 (#179), Grok 4.6: 44.5 (#66)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1289 | 1454 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Grok 4.6 leads
GPT-4o: 52.6 (#166), Grok 4.6: 62.3 (#80)
| Benchmark | GPT-4o | Grok 4.6 |
|---|---|---|
| LMArena Text | 1300 | 1428 |
| LMArena Creative Writing | 1292 | 1428 |
| LMArena Multi-Turn | 1302 | 1425 |
| Short-Story Creative Writing | 81.8% | — |
| WildBench | 82.8% | — |
| LiveBench Language | 47.6% | — |
Frequently asked questions
Is GPT-4o better than Grok 4.6?
Grok 4.6 is the stronger model overall, scoring 56.9 to 28.6 on the Noometry Index.
Which is cheaper, GPT-4o or Grok 4.6?
Grok 4.6 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-4o lists at $2.50 and $10.
Is GPT-4o or Grok 4.6 better for coding?
Grok 4.6 scores higher on coding benchmarks: 58.5 versus 24.8 in the Noometry coding category.
Which has the bigger context window?
Grok 4.6 does, with 500K tokens against 128K.
How many benchmarks do GPT-4o and Grok 4.6 share?
31 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and Grok 4.6 has 49.