Grok 4.7 ranks first on 2 of the 156 benchmarks it has scores on.
Closed weights Flagship
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.
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Longform Creative Writing (EQ-Bench) | 83.0 | #5 of 119 | Official leaderboard | |
| Creative Writing v3 | 79.8 | #13 of 120 | Official leaderboard | |
| SimpleQA Verified | 59.4% | #12 of 56 | Official leaderboard | |
| LMArena (Elo) | 1,441 | #76 of 333 | Official leaderboard | |
| Extended NYT Connections | 76.8% | #44 of 97 | Official leaderboard | |
| FACTS Grounding v2 | 68.6% | #8 of 15 | Official leaderboard | |
| FACTS Parametric V2 | 54.3% | #12 of 20 | Official leaderboard | |
| HLE-Diamond | 25% | #17 of 18 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| DTBench | 93.3% | #17 of 178 | Official leaderboard | |
| GPQA Diamond | 92.7% | #23 of 219 | Epoch AI | |
| OTIS Mock AIME 2024-2025 | 98.1% | #29 of 202 | Epoch AI | |
| CritPt | 18% | #36 of 160 | Epoch AI | |
| ARC-AGI-2 | 61.4% | #31 of 95 | Official leaderboard | |
| LatchBio Capabilities v1.0 | 47.4% | #9 of 27 | Official leaderboard | |
| ARC-AGI | 90.2% | #33 of 92 | Official leaderboard | |
| Terminal-Bench-Science | 14.3% | #8 of 17 | Official leaderboard | |
| FrontierPhysics | 6.50% | #9 of 19 | Official leaderboard | |
| ArXivMath | 54% | #3 of 6 | Official leaderboard | |
| ProofBench | 26% | #47 of 83 | Epoch AI | |
| FrontierMath Tiers 1-3 v2 | 53% | #52 of 89 | Epoch AI | |
| MysteryMechanism | 25.2% | #17 of 29 | Official leaderboard | |
| BrokenArXiv | 38.5% | #4 of 6 | Official leaderboard | |
| ProofBench v1.1 | 26% | #38 of 51 | Official leaderboard | |
| BixBench | 88.4% | #3 of 4 | Model card | |
| FrontierMath Tier 4 v2 | 17.1% | #55 of 71 | Epoch AI | |
| CritPt v2 | 50.6% | #9 of 11 | Official leaderboard | |
| Human Behavior Bench | 75.2 | #10 of 11 | Official leaderboard | |
| FrontierMath Erdős | 0% | #14 of 14 | Epoch AI |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vibe Code Bench | 86.2% | #13 of 115 | Official leaderboard | |
| WebDev Arena | 1,638 | #12 of 105 | Epoch AI | |
| Code Arena | 1,637 | #19 of 124 | Official leaderboard | |
| FrontierCode | 47.6% | #11 of 46 | Epoch AI | |
| Code Migration | 44.8% | #19 of 78 | Official leaderboard | |
| FrontierCode 2.0 | 28.6% | #8 of 26 | Official leaderboard | |
| DeepSWE | 71% | #16 of 51 | Model card | |
| Image-to-WebDev Arena | 1,586 | #19 of 55 | Official leaderboard | |
| OpenDesign Arena | 82.9 | #7 of 19 | Official leaderboard | |
| Code Migration v1.1 | 53.2% | #12 of 29 | Official leaderboard | |
| IOI v2 (Vals) | 57.7% | #19 of 45 | Official leaderboard | |
| Next.js Agent Evals | 94% | #6 of 14 | Official leaderboard | |
| SWE-Marathon | 46% | #5 of 11 | Model card | |
| CursorBench | 46.3% | #9 of 19 | Epoch AI | |
| Senior SWE-Bench | 27.4% | #12 of 23 | Official leaderboard | |
| FrontierSWE | 29.5% | #11 of 21 | Epoch AI | |
| SWE-Marathon v1.1 | 45.6% | #9 of 17 | Official leaderboard | |
| RSI-Exam | 38.4% | #10 of 18 | Official leaderboard | |
| ViBench 2.0 | 1.50% | #5 of 9 | Official leaderboard | |
| WeirdML v3 | 8.45% | #15 of 25 | Official leaderboard | |
| Vals RSI Index | 0.20 | #16 of 25 | Official leaderboard | |
| Brokk Power Ranking | 5.00 | #14 of 19 | Official leaderboard | |
| Web Clone Bench | 2.80% | #7 of 9 | Official leaderboard | |
| DeepRSI | 685 | #6 of 7 | Official leaderboard | |
| CursorBench 4.0 | 46.3% | #3 of 3 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Game Arena Ultimate Tic-Tac-Toe | 544 | #3 of 36 | Official leaderboard | |
| Hermes Bench | 74.4% | #2 of 17 | Official leaderboard | |
| Game Arena Nine Men's Morris | 908 | #6 of 36 | Official leaderboard | |
| Game Arena Lines of Action | 691 | #8 of 36 | Official leaderboard | |
| LLM Chess | 822 | #29 of 124 | Official leaderboard | |
| Kaggle Game Arena | 335 | #13 of 39 | Official leaderboard | |
| ARC-AGI-3 | 10.1% | #8 of 24 | Official leaderboard | |
| Game Arena Four in a Row | 471 | #13 of 38 | Official leaderboard | |
| Game Arena Clobber | 334 | #14 of 36 | Official leaderboard | |
| Agent Arena | 0.02 | #21 of 53 | Official leaderboard | |
| Game Arena Chess | 844 | #16 of 39 | Official leaderboard | |
| Mystery Game Puzzles | 29% | #33 of 80 | Epoch AI | |
| Game Arena Go | 1,007 | #15 of 36 | Official leaderboard | |
| Game Arena Word Association | 274 | #15 of 36 | Official leaderboard | |
| Chess Puzzles | 41% | #35 of 83 | Epoch AI | |
| Game Arena Coin Game | 376 | #15 of 35 | Official leaderboard | |
| Terminal-Bench 4.0 | 37.6% | #10 of 23 | Official leaderboard | |
| Game Arena Dots and Boxes | 927 | #17 of 37 | Official leaderboard | |
| Game Arena Word Art | 492 | #17 of 35 | Official leaderboard | |
| Game Arena Dark Hex | 255 | #19 of 38 | Official leaderboard | |
| WitnessBench | 14.8% | #6 of 12 | Official leaderboard | |
| Game Arena Reversi | 471 | #19 of 36 | Official leaderboard | |
| Game Arena Checkers | 879 | #24 of 38 | Official leaderboard | |
| Game Arena Bargaining | 170 | #23 of 36 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MedScribe | 89.4% | #7 of 98 | Official leaderboard | |
| Vending-Bench 2 | $10,537 | #9 of 66 | Official leaderboard | |
| Ivo-micro1 ContractBench | 58.3% | #1 of 7 | Official leaderboard | |
| Harvey LAB v1.1 (HLAB-Vals-1.1) | 18.3% | #5 of 33 | Official leaderboard | |
| Harvey LAB (Legal Agent Benchmark) | 12.5% | #13 of 80 | Official leaderboard | |
| Legal Research Bench | 47.1% | #15 of 88 | Official leaderboard | |
| Tax Agent Bench | 34.2% | #14 of 79 | Official leaderboard | |
| Vals Legal Index | 72.6% | #7 of 33 | Official leaderboard | |
| MedCode | 49.5% | #22 of 96 | Official leaderboard | |
| EEBench | 64% | #6 of 26 | Official leaderboard | |
| LegalBench | 84.4% | #34 of 138 | Official leaderboard | |
| Vending-Bench Arena | $7,930 | #8 of 32 | Official leaderboard | |
| Excel Modeling Benchmark | 67% | #20 of 76 | Official leaderboard | |
| Vals Finance Index | 58.5% | #11 of 35 | Official leaderboard | |
| PolicyBench | 87.6% | #16 of 50 | Official leaderboard | |
| Excel Modeling Benchmark v1.1 | 71.5% | #12 of 37 | Official leaderboard | |
| Short Form Bench | $2,578 | #3 of 9 | Official leaderboard | |
| Immigration Bench | 68% | #9 of 24 | Official leaderboard | |
| Physical-Engineering's Last Exam | 0.30 | #7 of 18 | Official leaderboard | |
| APEX-Agents | 54.6% | #22 of 56 | Epoch AI | |
| Public Benefits Bench | 65.6% | #21 of 50 | Official leaderboard | |
| PRBench | 47.6% | #19 of 44 | Official leaderboard | |
| Vals Finance Agent v2 | 52.3% | #36 of 80 | Official leaderboard | |
| Vals Index v2.2 | 58.6% | #15 of 32 | Official leaderboard | |
| Legora BAR | 102.0 | #8 of 17 | Official leaderboard | |
| CADGenBench | 44.4% | #1 of 2 | Model card | |
| Parametric CAD Bench | 34.8% | #6 of 11 | Official leaderboard | |
| Vals Finance Agent v2.1 | 60.6% | #21 of 35 | Official leaderboard | |
| RCM-Bench-1 | 88.2% | #12 of 20 | Official leaderboard | |
| Holdings-IndividualTaxBench | 72.3% | #8 of 13 | Official leaderboard | |
| RealmPathology | 73.1% | #8 of 13 | Official leaderboard | |
| LegalReasonBench | 40.4% | #5 of 8 | Official leaderboard | |
| Thrive Holdings helpdesk ticket resolution benchmark | 46.1% | #10 of 15 | Official leaderboard | |
| AA-Briefcase v1.1 | 1,657 | #4 of 6 | Model card | |
| Clipping Bench | $384 | #14 of 19 | Official leaderboard | |
| CAD Arena | 0.26 | #9 of 12 | Official leaderboard | |
| HealthBench Professional | 56.7% | #20 of 25 | Model card | |
| MissionReport | 59.5% | #8 of 9 | Official leaderboard | |
| RWE-bench | 0% | #10 of 11 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vision Arena | 1,250 | #46 of 136 | Official leaderboard | |
| Blueprint-Bench 2 | 32.5% | #16 of 35 | Official leaderboard | |
| GDP.pdf | 22.8% | #20 of 39 | Epoch AI | |
| GeoBench | 61% | #35 of 48 | Official leaderboard | |
| DataCenterBench Tier 1 | 13.3% | #8 of 9 | Official leaderboard | |
| Furniture Assembly Benchmark | 20.8% | #33 of 35 | Epoch AI | |
| 1H-VideoQA | 59.4% | #19 of 20 | Official leaderboard | |
| Dual-Arm Sim | 6.30 | #10 of 10 | Official leaderboard | |
| Robo-bench | 0.21% | #7 of 7 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BioSecBench-Refusal | 62.4% | #1 of 22 | Official leaderboard | |
| Virology Capabilities Test (VCT) | 63% | #3 of 28 | Model card | |
| TxBench-Antibody-Discovery | 52% | #7 of 28 | Official leaderboard | |
| CWE-bench | 68% | #3 of 12 | Official leaderboard | |
| VariantBench | 45.5% | #7 of 27 | Official leaderboard | |
| MetagenomicsBench | 48.7% | #7 of 26 | Official leaderboard | |
| BioSecBench-Function | 43.3% | #9 of 27 | Official leaderboard | |
| Standard jailbreaks (xAI) | 0.01% | #1 of 3 | Model card | |
| TxBench-Oligo-Discovery | 41.9% | #10 of 28 | Official leaderboard | |
| BioSecBench-Surveillance | 48% | #10 of 27 | Official leaderboard | |
| OncologyBench | 18.7% | #14 of 30 | Official leaderboard | |
| CyberGym | 80.3% | #16 of 33 | Model card | |
| Human Pathogen Capabilities Test v2 (HPCT v2) | 54.1% | #35 of 71 | Official leaderboard | |
| BioUseBench | 91.4% | #1 of 2 | Model card | |
| Long-horizon jailbreaks (xAI) | 0.65% | #1 of 2 | Model card | |
| CyberBench v1.1 (Vals) | 69.5% | #22 of 43 | Official leaderboard | |
| TxBench-Preclinical-Pharmacology | 54.7% | #16 of 28 | Official leaderboard | |
| SRE-Bench | 3.82% | #19 of 33 | Official leaderboard | |
| Autointent-Bio | 100% | #2 of 3 | Model card | |
| Autointent-Chem | 99.9% | #2 of 3 | Model card | |
| LAB-Bench | 76.8% | #2 of 3 | Model card | |
| xAI sycophancy eval | 0.03% | #2 of 3 | Model card | |
| EpiBench | 20.8% | #19 of 27 | Official leaderboard | |
| SpatialBench-Long | 24.4% | #20 of 27 | Official leaderboard | |
| scBench-Long | 11.4% | #22 of 27 | Official leaderboard | |
| DeepsecBench | 10.9 | #23 of 28 | Official leaderboard | |
| WellnessBench | 48% | #10 of 12 | Official leaderboard | |
| General refusals (xAI) | 1.10% | #6 of 7 | Model card | |
| CheatBench | 78% | #9 of 10 | Official leaderboard | |
| CVE-Bench | 36.6% | #13 of 14 | Model card | |
| DistressBench | 52.4% | #41 of 42 | Official leaderboard | |
| StrongREJECT | 2.00% | #22 of 22 | Model card | |
| MASK | 0% | #21 of 21 | Model card | |
| Child safety (xAI) | 0% | #3 of 3 | Model card | |
| FORTRESS-RN | 97.9% | #3 of 3 | Model card | |
| HackerBench | 4.02% | #3 of 3 | Model card | |
| ProtocolQA Open-Ended (xAI) | 70.4% | #3 of 3 | Model card | |
| Self-harm refusals (xAI) | 1.05% | #3 of 3 | Model card | |
| Biosecurity VCT | 41.5% | #2 of 2 | Model card | |
| CathedralBench | 29% | #1 of 1 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Conceptual Reasoning Index | 63.1 | #27 of 149 | Official leaderboard | |
| ACCoRD | 74.0 | #28 of 150 | Official leaderboard | |
| LMCA | 49.4% | #28 of 139 | Epoch AI | |
| SAGE v1.1 (Vals) | 88.7% | #7 of 28 | Official leaderboard | |
| Hermes Index | 39.3 | #6 of 17 | Official leaderboard | |
| LiveBench | 70.1% | #11 of 28 | Official leaderboard | |
| Vals Science Index | 34.8% | #15 of 23 | Official leaderboard | |
| SAGE (Vals) | 40.8% | #58 of 81 | Official leaderboard |
xAI's models in color, against every other model on the Epoch Capabilities Index, by release date.
The benchmarks the lab chose to headline, with the scores it gave.
Every result in the longer documents the lab published with it. 29 of them aren't in the launch post.