Grok 4
| Released | 2025-07-09 |
|---|---|
| Price per 1M | Not listed |
| Overall score | 50.7 · rank 105 |
| Provider | xAI |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 50.7 | 105 | 22 |
| Overall: Coding | 51.4 | 101 | 13 |
| Overall: Reasoning | 52.9 | 93 | 9 |
| Overall: Agentic work | 50.4 | 112 | 13 |
| Overall: Frontend | 49.7 | 90 | 6 |
| Coding | 51.8 | 106 | 3 |
| Agents | 48.7 | 112 | 5 |
| Reasoning | 54.9 | 84 | 5 |
| Math | 51.3 | 102 | 3 |
| Knowledge | 51.5 | 86 | 1 |
| General | 47.9 | 128 | 2 |
| Multimodal | 43.6 | 96 | 3 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 50.7 ±1.3 | 22 |
| High | 51.1 ±5.0 | 1 |
| Thinking | 54.0 ±5.7 | 2 |
Evidence
Coding 51.8 ±2.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Aider Polyglot | 79.6% | Epoch AI | 2026-11-11 |
| SWE-bench Verified (Vals run) | 57.8% | Vals AI | 2026-09-01 |
| LiveCodeBench (Vals run) | 83.2% | Vals AI | 2026-09-01 |
Agents 48.7 ±2.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 2.0 | 27.2% | Epoch AI | 2025-11-02 |
| METR task suite | 66.6% | Epoch AI | 2026-11-11 |
| BFCL v4 | 63.0% | BFCL | 2026-04-12 |
| DeepResearch Bench | 47.3% | Epoch AI | 2026-11-11 |
| BALROG | 43.6% | Epoch AI | 2026-11-11 |
Reasoning 54.9 ±2.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 16.0% | Epoch AI | 2026-11-11 |
| ARC-AGI-2 (ARC Prize verified) | 16.0% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 66.7% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 66.7% | Epoch AI | 2026-11-11 |
| GPQA Diamond | 87.0% | Epoch AI | 2026-11-11 |
| GPQA Diamond (Vals run) | 88.1% | Vals AI | 2026-09-01 |
| Kagi Reasoning Benchmark | 73.6% | Kagi | 2026-11-11 |
| SimpleBench | 60.5% | SimpleBench | 2026-11-11 |
| SimpleBench | 60.5% | Epoch AI | 2026-11-11 |
Math 51.3 ±2.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 (2025 set) | 19.7% | Epoch AI | 2025-11-13 |
| FrontierMath Tier 4 (2025 set) | 2.1% | Epoch AI | 2025-08-11 |
| OTIS Mock AIME 2024-2025 | 84.0% | Epoch AI | 2026-11-11 |
Knowledge 51.5 ±4.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMLU-Pro (Vals run) | 85.3% | Vals AI | 2026-09-01 |
General 47.9 ±3.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,411 | LMArena | 2026-11-09 |
| EQ-Bench Longform Creative Writing | 45.7% | EQ-Bench | 2026-11-11 |
Multimodal 43.6 ±9.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Vision | 1,184 | LMArena | 2026-11-09 |
| MMMU-Pro (Vals run) | 76.3% | Vals AI | 2026-09-01 |
| GeoBench | 45.0% | Epoch AI | 2026-11-11 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| Aider Polyglot | 79.6% | Epoch AI | A stronger configuration result was used |
| BFCL v4 | 61.4% | BFCL | A stronger configuration result was used |
| Terminal-Bench 2.0 | 25.4% | Epoch AI | A stronger configuration result was used |
| Terminal-Bench 2.0 | 23.2% | Epoch AI | A stronger configuration result was used |
Names matched to this model
grok-4, grok-4-0709