Model comparison
GPT-5 vs Grok 4.7
Grok 4.7 is the stronger model overall, scoring 53.1 to 50.9 on the Noometry Index.
Last verified . 30 shared benchmarks.
Summary
- They share 30 benchmarks with published results for both. GPT-5 scores higher in 3 categories and Grok 4.7 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 43.1.
- The biggest single-benchmark swing is ProofBench: 18% for GPT-5 and 34% for Grok 4.7.
- Grok 4.7 is cheaper at $2 / $6 per million input/output tokens, against $1.25 / $10 for GPT-5.
- Grok 4.7 accepts more context: 500K tokens versus 400K.
Side by side
| GPT-5 | Grok 4.7 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 50.9 | 53.1 |
| Released | 2025-08-07 | 2026-09-21 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $1.25 | $2 |
| Output $ / M tokens | $10 | $6 |
| Results tracked | 69 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.7 leads
GPT-5: 50.3 (#47), Grok 4.7: 58.0 (#18)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| LMArena WebDev | 1418 | 1639 |
| SciCode | 42.9% | 57.8% |
| LMArena Coding | 1436 | 1427 |
| SWE-bench Verified | 73.6% | — |
| FrontierCode | — | 47.6% |
| SWE-bench Verified (bash only) | 65% | — |
| Aider Polyglot | 88% | — |
| CursorBench | — | 46.3% |
| FrontierSWE | — | 29.5% |
| GSO | 6.9% | — |
| WeirdML | 60.7% | — |
| ALE-Bench | 1,162 | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use Grok 4.7 leads
GPT-5: 33.1 (#56), Grok 4.7: 36.7 (#37)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| Terminal-Bench | 49.6% | — |
| APEX-Agents | — | 54.6% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| DeepResearch Bench | 49.6% | — |
| BALROG | 32.8% | — |
| GDP.pdf | — | 22.8% |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
| Vending-Bench 2 | — | 10,537 |
Reasoning Grok 4.7 leads
GPT-5: 38.3 (#64), Grok 4.7: 49.1 (#40)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| CritPt | 12.6% | 18% |
| Chess Puzzles | 37% | 38% |
| LMArena Hard Prompts | 1416 | 1413 |
| Mystery Game Puzzles | 23% | 29% |
| DTBench | 90.7% | 96% |
| LMCA | 40% | 49.4% |
| Epoch Capabilities Index | 150 | 153.53 |
| ARC-AGI-2 | 9.9% | — |
| SimpleBench | 56.7% | — |
| Kagi LLM Benchmark | 72.7% | — |
| NYT Connections (extended) | — | 76.8% |
| ARC-AGI-1 | 65.7% | — |
| EnigmaEval | 10.5% | — |
| EBR-Bench | 12.7% | — |
| ForecastBench | 61.4 | — |
Math Grok 4.7 leads
GPT-5: 55.0 (#44), Grok 4.7: 57.8 (#39)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 53% |
| FrontierMath Tier 4 | 22% | 17.1% |
| OTIS Mock AIME 2024-2025 | 91.4% | 98.1% |
| ProofBench | 18% | 34% |
| LMArena Math | 1407 | 1407 |
| Omni-MATH | 64.7% | — |
| MATH Level 5 | 98.1% | — |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |
Knowledge Grok 4.7 leads
GPT-5: 56.6 (#43), Grok 4.7: 62.8 (#22)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| GPQA Diamond | 86.2% | 92.7% |
| SimpleQA Verified | 50.1% | 56% |
| LMArena Expert | 1419 | 1422 |
| Humanity's Last Exam | 25.3% | — |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.3% | — |
| Vectara Hallucination Rate | 14.7% | — |
| GPQA (HELM) | 79.2% | — |
Multimodal GPT-5 leads
GPT-5: 46.8 (#13), Grok 4.7: 35.5 (#87)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| LMArena Vision | 1232 | 1228 |
| GeoBench | 81% | — |
| VPCT | 66% | — |
| Blueprint-Bench 2 | — | 32.5% |
| Furniture Assembly | — | 20.8% |
Multilingual Too close to call
GPT-5: 51.4 (#110), Grok 4.7: 50.8 (#116)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| LMArena Non-English | 1397 | 1389 |
| LMArena Chinese | 1422 | 1455 |
| LMArena French | 1410 | 1455 |
| LMArena Russian | 1406 | 1397 |
| LMArena Spanish | 1399 | 1400 |
| LMArena German | 1416 | — |
| LMArena Japanese | 1409 | — |
| LMArena Korean | 1360 | — |
Instruction Following Too close to call
GPT-5: 73.8 (#113), Grok 4.7: 74.1 (#105)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| LMArena Instruction Following | 1388 | 1404 |
| IFEval | 87.5% | — |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Grok 4.7: 43.1 (#104)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| LMArena Longer Query | 1399 | 1413 |
| Fiction.LiveBench | 97.2% | — |
Writing & Preference Grok 4.7 leads
GPT-5: 63.4 (#65), Grok 4.7: 70.0 (#24)
| Benchmark | GPT-5 | Grok 4.7 |
|---|---|---|
| LMArena Text | 1406 | 1399 |
| LMArena Creative Writing | 1365 | 1391 |
| EQ-Bench Creative Writing | 1627 | 2007 |
| LMArena Multi-Turn | 1426 | 1393 |
| Short-Story Creative Writing | 86% | — |
| WildBench | 85.7% | — |
Frequently asked questions
Is GPT-5 better than Grok 4.7?
Grok 4.7 is the stronger model overall, scoring 53.1 to 50.9 on the Noometry Index.
Which is cheaper, GPT-5 or Grok 4.7?
Grok 4.7 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GPT-5 or Grok 4.7 better for coding?
Grok 4.7 scores higher on coding benchmarks: 58.0 versus 50.3 in the Noometry coding category.
Which has the bigger context window?
Grok 4.7 does, with 500K tokens against 400K.
How many benchmarks do GPT-5 and Grok 4.7 share?
30 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Grok 4.7 has 39.