Grok 4.3
Grok 4.3 ranks first on CaseLaw v2.
74
Benchmarks with a score
1
Ranked first, out of five or more models
1
In the top three
xAI Apr 2026
Grok 4.3
Closed weights Flagship
No Epoch Capabilities Index score yet
- ZeroBench 13% #23 /66
- Harvey LAB (Legal Agent Benchmark) 0.42% #62 /80
- SkillsBench 41.7% #16 /21
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BullshitBench | 61.8% | #17 of 107 | Official leaderboard | |
| LMArena (Elo) | 1,442 | #73 of 333 | Official leaderboard | |
| Wer wird Millionär? LLM Benchmark | 76200.0 | #22 of 76 | Official leaderboard | |
| BullshitBench V2 | 50% | #43 of 122 | Official leaderboard | |
| Phare | 68.6% | #26 of 62 | Official leaderboard | |
| Extended NYT Connections | 55.2% | #62 of 97 | Official leaderboard | |
| Judgemark v4 | 49.6 | #26 of 40 | Official leaderboard | |
| EQ-Bench 4 | 1,065 | #29 of 32 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| DTBench | 84.5% | #45 of 178 | Official leaderboard | |
| ForecastBench | 63.4 | #53 of 86 | Official leaderboard | |
| ProofBench | 11% | #42 of 59 | Official leaderboard | |
| Leipzig Benchmark | 6.00% | #18 of 19 | Official leaderboard | |
| ResearchClawBench | 12.4 | #23 of 23 | Official leaderboard |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| WeirdML | 49.9% | #47 of 112 | Official leaderboard | |
| IOI (Vals) | 15.3% | #28 of 55 | Official leaderboard | |
| Vibe Code Bench | 19.4% | #85 of 115 | Official leaderboard | |
| Code Arena | 1,358 | #96 of 124 | Official leaderboard | |
| Code Migration | 6.79% | #68 of 78 | Official leaderboard | |
| Image-to-WebDev Arena | 1,375 | #52 of 55 | Official leaderboard |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LLM Chess | 179 | #56 of 124 | Official leaderboard | |
| PinchBench 2.0 | 73.7% | #28 of 57 | Official leaderboard | |
| Pencil Puzzle Bench | 3.30% | #29 of 53 | Official leaderboard | |
| Game Arena Coin Game | 279 | #23 of 35 | Official leaderboard | |
| LMArena Search Arena | 1,193 | #23 of 35 | Official leaderboard | |
| PACT | 1,504 | #18 of 26 | Official leaderboard | |
| Game Arena Word Association | 110 | #27 of 36 | Official leaderboard | |
| Game Arena Five in a Row | 367 | #15 of 20 | Official leaderboard | |
| SkillsBench | 41.7% | #16 of 21 | Official leaderboard | |
| Game Arena Chess | 591 | #30 of 39 | Official leaderboard | |
| Game Arena Four in a Row | 197 | #30 of 38 | Official leaderboard | |
| Game Arena Poker | -16.53 | #25 of 31 | Official leaderboard | |
| Game Arena Lines of Action | 293 | #32 of 36 | Official leaderboard | |
| Game Arena Ultimate Tic-Tac-Toe | 27 | #32 of 36 | Official leaderboard | |
| Game Arena Bargaining | 131 | #33 of 36 | Official leaderboard | |
| Game Arena Dots and Boxes | 442 | #34 of 37 | Official leaderboard | |
| Game Arena Dark Hex | 70 | #35 of 38 | Official leaderboard | |
| Kaggle Game Arena | 145 | #36 of 39 | Official leaderboard | |
| Game Arena Go | 130 | #34 of 36 | Official leaderboard | |
| Game Arena Word Art | 2 | #34 of 35 | Official leaderboard | |
| Game Arena Nine Men's Morris | 12 | #35 of 36 | Official leaderboard | |
| Game Arena Reversi | 17 | #35 of 36 | Official leaderboard | |
| Game Arena Checkers | 0 | #38 of 38 | Official leaderboard | |
| Game Arena Clobber | 0 | #36 of 36 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| CaseLaw v2 | 79.3% | #1 of 53 | Official leaderboard | |
| CorpFin v2 | 68.5% | #8 of 119 | Official leaderboard | |
| LegalBench | 84.5% | #32 of 138 | Official leaderboard | |
| Vals Finance Agent | 53.8% | #18 of 49 | Official leaderboard | |
| TaxEval v2 | 70.8% | #81 of 129 | Official leaderboard | |
| PolicyBench | 80.3% | #34 of 50 | Official leaderboard | |
| MedCode | 38.1% | #68 of 96 | Official leaderboard | |
| ProfBench | 43.3% | #53 of 72 | Official leaderboard | |
| MedScribe | 74.4% | #75 of 98 | Official leaderboard | |
| Harvey LAB (Legal Agent Benchmark) | 0.42% | #62 of 80 | Official leaderboard | |
| Legal Research Bench | 15.4% | #70 of 88 | Official leaderboard | |
| Vals Finance Agent v2 | 37.7% | #66 of 80 | Official leaderboard | |
| MortgageTax | 48.3% | #76 of 86 | Official leaderboard | |
| Public Benefits Bench | 51.7% | #45 of 50 | Official leaderboard | |
| Tax Agent Bench | 6.18% | #73 of 79 | Official leaderboard | |
| Vending-Bench 2 | $35 | #61 of 66 | Official leaderboard | |
| Excel Modeling Benchmark | 18.4% | #71 of 76 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| ZeroBench | 13% | #23 of 66 | Official leaderboard | |
| Vision Arena | 1,242 | #54 of 136 | Official leaderboard | |
| GeoBench | 56% | #39 of 48 | Official leaderboard | |
| Blueprint-Bench 2 | 0% | #35 of 35 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| PCT (Polarized Contrastive Pairs) | 51.6% | #35 of 46 | Official leaderboard | |
| Dictatorship Eval | 37.9% | #16 of 20 | Official leaderboard | |
| CyberBench (Vals) | 47.5% | #22 of 26 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Conceptual Reasoning Index | 52.9 | #58 of 149 | Official leaderboard | |
| ACCoRD | 65.3 | #71 of 150 | Official leaderboard | |
| CAIS Vision Capabilities Index | 48.4 | #29 of 53 | Official leaderboard | |
| CAIS Risk Index | 54.5 | #40 of 65 | Official leaderboard | |
| CAIS Text Capabilities Index | 23.4 | #45 of 66 | Official leaderboard | |
| Vals Multimodal Index | 43.3% | #29 of 33 | Official leaderboard | |
| SAGE (Vals) | 19.7% | #81 of 81 | Official leaderboard |