Model comparison
GPT-5 vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 50.9 on the Noometry Index.
Last verified . 39 shared benchmarks.
Summary
- They share 39 benchmarks with published results for both. GPT-5 scores higher in 2 categories and Grok 4.5 in 8 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 44.8.
- The biggest single-benchmark swing is ARC-AGI-2: 9.9% for GPT-5 and 52.6% for Grok 4.5.
- Grok 4.5 is cheaper at $2 / $6 per million input/output tokens, against $1.25 / $10 for GPT-5.
- Grok 4.5 accepts more context: 500K tokens versus 400K.
Side by side
| GPT-5 | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 50.9 | 55.0 |
| Released | 2025-08-07 | 2026-07-08 |
| 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 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
GPT-5: 50.3 (#47), Grok 4.5: 52.2 (#35)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena WebDev | 1418 | 1553 |
| SciCode | 42.9% | 54.1% |
| WeirdML | 60.7% | 46.4% |
| LMArena Coding | 1436 | 1474 |
| ALE-Bench | 1,162 | 1,309 |
| SWE-bench Verified | 73.6% | — |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| SWE-bench Verified (bash only) | 65% | — |
| Aider Polyglot | 88% | — |
| GSO | 6.9% | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use Grok 4.5 leads
GPT-5: 33.1 (#56), Grok 4.5: 44.4 (#17)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena Search | 1133 | 1213 |
| Terminal-Bench | 49.6% | — |
| APEX-Agents | — | 56.2% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| τ²-bench Banking | — | 47.9% |
| DeepResearch Bench | 49.6% | — |
| PostTrainBench | — | 23.4% |
| BALROG | 32.8% | — |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| METR Time Horizons | 69.6% | — |
| Vending-Bench 2 | — | 3,887 |
Reasoning Grok 4.5 leads
GPT-5: 38.3 (#64), Grok 4.5: 56.1 (#25)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 9.9% | 52.6% |
| SimpleBench | 56.7% | 70% |
| Kagi LLM Benchmark | 72.7% | 83.5% |
| ARC-AGI-1 | 65.7% | 87.2% |
| CritPt | 12.6% | 15.4% |
| Chess Puzzles | 37% | 36% |
| LMArena Hard Prompts | 1416 | 1462 |
| DTBench | 90.7% | 96.5% |
| LMCA | 40% | 45.2% |
| Epoch Capabilities Index | 150 | 153.92 |
| NYT Connections (extended) | — | 79.9% |
| EnigmaEval | 10.5% | — |
| EBR-Bench | 12.7% | — |
| Mystery Game Puzzles | 23% | — |
| Surface Evolver Bench | — | 74.4% |
| ForecastBench | 61.4 | — |
Math Grok 4.5 leads
GPT-5: 55.0 (#44), Grok 4.5: 60.9 (#35)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 57.2% |
| FrontierMath Tier 4 | 22% | 24.4% |
| OTIS Mock AIME 2024-2025 | 91.4% | 97.8% |
| ProofBench | 18% | 31% |
| LMArena Math | 1407 | 1459 |
| Omni-MATH | 64.7% | — |
| MATH Level 5 | 98.1% | — |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |
Knowledge Grok 4.5 leads
GPT-5: 56.6 (#43), Grok 4.5: 62.3 (#24)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 86.2% | 93.4% |
| SimpleQA Verified | 50.1% | 48.3% |
| LMArena Expert | 1419 | 1466 |
| 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 4.5: 37.6 (#72)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena Vision | 1232 | 1288 |
| GeoBench | 81% | — |
| VPCT | 66% | — |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.5 leads
GPT-5: 51.4 (#110), Grok 4.5: 54.4 (#42)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1397 | 1440 |
| LMArena Chinese | 1422 | 1496 |
| LMArena French | 1410 | 1456 |
| LMArena German | 1416 | 1446 |
| LMArena Japanese | 1409 | 1428 |
| LMArena Korean | 1360 | 1404 |
| LMArena Russian | 1406 | 1448 |
| LMArena Spanish | 1399 | 1450 |
Instruction Following Grok 4.5 leads
GPT-5: 73.8 (#113), Grok 4.5: 76.0 (#48)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1388 | 1446 |
| IFEval | 87.5% | — |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Grok 4.5: 44.8 (#56)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1399 | 1463 |
| Fiction.LiveBench | 97.2% | — |
Writing & Preference Grok 4.5 leads
GPT-5: 63.4 (#65), Grok 4.5: 65.8 (#42)
| Benchmark | GPT-5 | Grok 4.5 |
|---|---|---|
| LMArena Text | 1406 | 1448 |
| LMArena Creative Writing | 1365 | 1442 |
| EQ-Bench Creative Writing | 1627 | 1579 |
| LMArena Multi-Turn | 1426 | 1456 |
| Short-Story Creative Writing | 86% | — |
| WildBench | 85.7% | — |
Frequently asked questions
Is GPT-5 better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 50.9 on the Noometry Index.
Which is cheaper, GPT-5 or Grok 4.5?
Grok 4.5 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 4.5 better for coding?
Grok 4.5 scores higher on coding benchmarks: 52.2 versus 50.3 in the Noometry coding category.
Which has the bigger context window?
Grok 4.5 does, with 500K tokens against 400K.
How many benchmarks do GPT-5 and Grok 4.5 share?
39 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Grok 4.5 has 52.