Model comparison
GPT-4 vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 28.9 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. GPT-4 scores higher in 0 categories and Grok 4.5 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4.5 leads 60.7 to 10.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.1% for GPT-4 and 97.8% for Grok 4.5.
- Grok 4.5 is cheaper at $2 / $6 per million input/output tokens, against $30 / $60 for GPT-4.
- Grok 4.5 accepts more context: 500K tokens versus 8K.
Side by side
| GPT-4 | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 28.9 | 55.0 |
| Released | 2023-03-14 | 2026-07-08 |
| Weights | Proprietary | Proprietary |
| Context window | 8K | 500K |
| Max output | 8K | 500K |
| Input $ / M tokens | $30 | $2 |
| Output $ / M tokens | $60 | $6 |
| Results tracked | 38 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
GPT-4: 31.2 (#277), Grok 4.5: 51.7 (#34)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| WeirdML | 12.4% | 46.4% |
| LMArena Coding | 1254 | 1473 |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| LMArena WebDev | — | 1552 |
| SciCode | — | 54.1% |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| ALE-Bench | — | 1,309 |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, Grok 4.5: 44.4 (#15)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| APEX-Agents | — | 56.2% |
| τ²-bench Banking | — | 47.9% |
| PostTrainBench | — | 23.4% |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| LMArena Search | — | 1213 |
| METR Time Horizons | 36.1% | — |
| Vending-Bench 2 | — | 3,887 |
Reasoning Grok 4.5 leads
GPT-4: 17.3 (#285), Grok 4.5: 56.0 (#23)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| Chess Puzzles | 4% | 36% |
| LMArena Hard Prompts | 1241 | 1462 |
| DTBench | 62.7% | 96.5% |
| LMCA | 17.1% | 45.2% |
| Epoch Capabilities Index | 125.88 | 154.02 |
| ARC-AGI-2 | — | 52.6% |
| SimpleBench | — | 70% |
| Kagi LLM Benchmark | — | 83.5% |
| NYT Connections (extended) | — | 79.9% |
| ARC-AGI-1 | — | 87.2% |
| CritPt | — | 15.4% |
| Mystery Game Puzzles | 12% | — |
| Surface Evolver Bench | — | 74.4% |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math Grok 4.5 leads
GPT-4: 10.1 (#304), Grok 4.5: 60.7 (#29)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 97.8% |
| LMArena Math | 1269 | 1457 |
| FrontierMath (Tiers 1-3) | — | 57.2% |
| FrontierMath Tier 4 | — | 24.4% |
| ProofBench | — | 31% |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge Grok 4.5 leads
GPT-4: 18.4 (#277), Grok 4.5: 62.3 (#21)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 35.7% | 93.4% |
| LMArena Expert | 1211 | 1466 |
| SimpleQA Verified | — | 48.3% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multimodal Not comparable
GPT-4: —, Grok 4.5: 37.8 (#63)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| LMArena Vision | — | 1287 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.5 leads
GPT-4: 40.8 (#210), Grok 4.5: 54.5 (#43)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1246 | 1439 |
| LMArena Chinese | 1242 | 1497 |
| LMArena French | 1282 | 1458 |
| LMArena German | 1251 | 1448 |
| LMArena Japanese | 1209 | 1425 |
| LMArena Korean | 1184 | 1402 |
| LMArena Russian | 1251 | 1447 |
| LMArena Spanish | 1261 | 1452 |
Instruction Following Grok 4.5 leads
GPT-4: 65.5 (#217), Grok 4.5: 76.3 (#45)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1241 | 1446 |
Long Context Grok 4.5 leads
GPT-4: 37.7 (#207), Grok 4.5: 44.8 (#54)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1244 | 1463 |
Writing & Preference Grok 4.5 leads
GPT-4: 35.0 (#263), Grok 4.5: 65.9 (#40)
| Benchmark | GPT-4 | Grok 4.5 |
|---|---|---|
| LMArena Text | 1263 | 1448 |
| LMArena Creative Writing | 1244 | 1442 |
| EQ-Bench Creative Writing | 752 | 1579 |
| LMArena Multi-Turn | 1257 | 1457 |
Frequently asked questions
Is GPT-4 better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 28.9 on the Noometry Index.
Which is cheaper, GPT-4 or Grok 4.5?
Grok 4.5 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-4 lists at $30 and $60.
Is GPT-4 or Grok 4.5 better for coding?
Grok 4.5 scores higher on coding benchmarks: 51.7 versus 31.2 in the Noometry coding category.
Which has the bigger context window?
Grok 4.5 does, with 500K tokens against 8K.
How many benchmarks do GPT-4 and Grok 4.5 share?
25 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Grok 4.5 has 52.