xAI
availableShows if the model has enough results for an index.Grok 4.5
Grok 4.5 is a reasoning model from xAI. 43 benchmarks count toward its score, in 7 categories.
IndexOverall score out of 100.66.0 ±3.0
CoverageShare of the index weight with results.95%
SpeedOutput tokens per second.49/s
Input / 1MUS dollars per 1M input tokens.$2
Output / 1MUS dollars per 1M output tokens.$6
ContextMaximum tokens in one request.500K
EloLMArena rating and rank.1450 (#45)
The index is a score out of 100. The ± range shows how much it can change.
30,103 votes. Elo shows what people prefer. It does not change the score.
CapabilitiesScore per category, out of 100.
Out of 100Results
43 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result as a score out of 100. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| OTIS Mock AIME 2024-2025 | Math | 97.8% | 66.2 | high effort | — | Epoch AI |
| GPQA diamond | Knowledge | 93.4% | 64.5 | high effort | — | Epoch AI |
| Artificial Analysis GPQA Diamond | Knowledge | 93.1% | 64.0 | — | — | Artificial Analysis |
| GPQA Diamond | Knowledge | 92.9% | 64.0 | — | 1 Sept 2026 | Vals AI |
| LiveBench Mathematics | Math | 90.8% | 69.3 | — | 25 Jun 2026 | LiveBench |
| VulcanBench v3 | Coding | 89.9% | 77.4 | — | — | VulcanBench contributors |
| MMLU Pro | Knowledge | 89.2% | 61.1 | — | 1 Sept 2026 | Vals AI |
| LiveCodeBench | Coding | 87.4% | 63.7 | — | 1 Sept 2026 | Vals AI |
| LiveBench Reasoning | Reasoning | 87.2% | 77.1 | — | 25 Jun 2026 | LiveBench |
| SWE-bench | Coding | 86.6% | 67.0 | — | 1 Sept 2026 | Vals AI |
| ARC-AGI-1 (semi-private) | Reasoning | 85.7% | 67.7 | high effort | — | ARC Prize Foundation |
| LiveBench Language | Knowledge | 82.8% | 70.4 | — | 25 Jun 2026 | LiveBench |
| Artificial Analysis MMMU-Pro | Multimodal | 80.4% | 65.3 | — | — | Artificial Analysis |
| Artificial Analysis Long Context Reasoning | Reasoning | 79.3% | 63.1 | — | — | Artificial Analysis |
| LiveBench Data Analysis | Reasoning | 73.0% | 57.4 | — | 25 Jun 2026 | LiveBench |
| Artificial Analysis Coding Index | Coding | 72.5% | 70.0 | — | — | Artificial Analysis |
| LiveBench Instruction Following | Instruction | 71.5% | 72.9 | — | 25 Jun 2026 | LiveBench |
| Vibe Code Bench v1.1 | Coding | 69.0% | 70.9 | OpenHands | 21 Sept 2026 | Vals AI |
| LiveBench Coding | Coding | 68.6% | 51.6 | — | 25 Jun 2026 | LiveBench |
| Terminal-Bench 2.1 | Agentic | 67.8% | 64.1 | — | 21 Sept 2026 | Vals AI |
| cursorBench32 | Coding | 66.7% | 74.8 | — | — | Benchmark authors |
| SkillsBench | Coding | 66.0% | 79.0 | OpenHands | 11 Sept 2026 | Vals AI |
| SWE-bench Pro | Coding | 64.7% | 66.6 | — | — | Xiang Deng et al. |
| MMMU Pro | Multimodal | 61.8% | 27.9 | — | 1 Sept 2026 | Vals AI |
| FrontierMath-Tiers-1-3-v2-Private | Math | 57.2% | 63.6 | high effort | — | Epoch AI |
| LiveBench Agentic Coding | Agentic | 56.5% | 69.9 | — | 25 Jun 2026 | LiveBench |
| Artificial Analysis SciCode | Coding | 55.0% | 69.2 | — | — | Artificial Analysis |
| DeepSWE | Agentic | 53.0% | 62.3 | — | — | Datacurve AI |
| ARC-AGI-2 (semi-private) | Reasoning | 52.6% | 66.3 | high effort | — | ARC Prize Foundation |
| Artificial Analysis Omniscience Accuracy | Knowledge | 51.6% | 77.7 | — | — | Artificial Analysis |
| SimpleQA Verified | Knowledge | 48.3% | 66.3 | high effort | — | Epoch AI |
| τ²-bench Banking | Agentic | 47.9% | 33.2 | high effort · Sierra | 4 Aug 2026 | Sierra Research |
| GDPval-AA normalized | Agentic | 43.5% | 69.0 | — | — | Artificial Analysis |
| Artificial Analysis Humanity's Last Exam | Knowledge | 42.7% | 70.9 | — | — | Artificial Analysis |
| Artificial Analysis Agentic Index | Agentic | 42.1% | 71.5 | — | — | Artificial Analysis |
| IOI | Coding | 40.6% | 63.3 | — | 21 Sept 2026 | Vals AI |
| Artificial Analysis Intelligence Index | Knowledge | 38.8% | 70.9 | — | — | Artificial Analysis |
| Code Migration | Coding | 36.6% | 68.3 | — | 21 Sept 2026 | Vals AI |
| Chess Puzzles | Reasoning | 36.0% | 70.1 | high effort | — | Epoch AI |
| ProofBench v1.1 | Math | 31.0% | 62.0 | — | 21 Sept 2026 | Vals AI |
| FrontierMath-Tier-4-v2-Private | Math | 24.4% | 59.6 | high effort | — | Epoch AI |
| Terminal-Bench 3.0 | Agentic | 15.7% | 66.8 | — | — | Ryan Marten et al. |
| Critical Physics Tasks | Reasoning | 15.4% | 70.7 | — | — | Artificial Analysis |
| Terminal-Bench 4.0.0 | Agentic | 12.4% | 68.2 | high effort · Grok Build | 21 Sept 2026 | Terminal-Bench |
| Agent Arena command recovery | Agentic | 6.6 | 74.8 | — | 15 Sept 2026 | LMArena |
| Terminal-Bench 4.0 | Agentic | 6.6% | 64.1 | — | 21 Sept 2026 | Vals AI |
| Agent Arena steerability | Agentic | 5.9 | 74.0 | — | 15 Sept 2026 | LMArena |
| Agent Arena task outcome | Agentic | 1.9 | 69.5 | — | 15 Sept 2026 | LMArena |
| ARC-AGI-3 (semi-private) | Reasoning | 0.3% | — | high effort | — | ARC Prize Foundation |
| ProgramBench | Coding | 0.0% | — | — | 21 Sept 2026 | Vals AI |
43 benchmarks count, from 48 of 50 results. A grey row does not count. Too few models took that benchmark.