Gemini 3.8 Flash ranks first on 5 of the 203 benchmarks it has scores on.
Google DeepMind Sep 2026 Closed weights Mid-size
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 | |
|---|---|---|---|---|
| FACTS Multimodal | 50.7% | #1 of 39 | Official leaderboard | |
| LMArena (Elo) | 1,496 | #9 of 333 | Official leaderboard | |
| SEA-HELM | 77.3 | #2 of 62 | Official leaderboard | |
| Extended NYT Connections | 97.4% | #5 of 97 | Official leaderboard | |
| SimpleQA Verified | 74.6% | #4 of 56 | Official leaderboard | |
| FACTS Benchmark Suite | 68.5% | #3 of 40 | Official leaderboard | |
| Humanity's Last Exam | 44.5% | #5 of 47 | Official leaderboard | |
| Global MMLU Lite | 95.3% | #5 of 43 | Official leaderboard | |
| FACTS Parametric | 77.3% | #7 of 45 | Official leaderboard | |
| Creative Writing v3 | 78.0 | #22 of 120 | Official leaderboard | |
| FACTS Parametric V2 | 78.7% | #4 of 20 | Official leaderboard | |
| Longform Creative Writing (EQ-Bench) | 76.8 | #25 of 119 | Official leaderboard | |
| HLE-Verified | 54.9% | #2 of 4 | Model card | |
| FACTS Grounding v2 | 68.2% | #9 of 15 | Official leaderboard | |
| HLE-Diamond | 33.3% | #13 of 18 | Official leaderboard | |
| FACTS Search V2 | 76.3% | #17 of 21 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GPQA Diamond | 95.4% | #4 of 219 | Epoch AI | |
| ARC-AGI | 98.5% | #2 of 92 | Official leaderboard | |
| SimpleBench | 82.4% | #6 of 100 | Official leaderboard | |
| DTBench | 92.9% | #18 of 178 | Official leaderboard | |
| OTIS Mock AIME 2024-2025 | 98.9% | #21 of 202 | Epoch AI | |
| ARC-AGI-2 | 89.2% | #10 of 95 | Official leaderboard | |
| CritPt | 18.3% | #34 of 160 | Epoch AI | |
| BioMysteryBench | 88.8% | #2 of 8 | Model card | |
| Surface Evolver Bench | 76.9% | #7 of 26 | Epoch AI | |
| MysteryMechanism | 36.5% | #8 of 29 | Official leaderboard | |
| Prophet Arena | 85.6 | #6 of 18 | Official leaderboard | |
| xbench-ScienceQA | 72.4% | #10 of 29 | Official leaderboard | |
| FrontierMath Tiers 1-3 v2 | 68.4% | #31 of 89 | Epoch AI | |
| ExplorationBench | 79% | #4 of 10 | Official leaderboard | |
| ProofBench | 48% | #35 of 83 | Epoch AI | |
| Leipzig Benchmark | 41% | #10 of 19 | Official leaderboard | |
| Terminal-Bench-Science | 12.4% | #9 of 17 | Official leaderboard | |
| Human Behavior Bench | 79.2 | #6 of 11 | Official leaderboard | |
| Verifying the Verifiers (Project APE CRED benchmark) | 79.5% | #14 of 23 | Official leaderboard | |
| ProofBench v1.1 | 48% | #32 of 51 | Official leaderboard | |
| ArXivMath | 52.6% | #4 of 6 | Official leaderboard | |
| LABBench2 | 86.2% | #2 of 3 | Model card | |
| FrontierMath Tier 4 v2 | 21.9% | #49 of 71 | Epoch AI | |
| LatchBio Capabilities v1.0 | 37.7% | #19 of 27 | Official leaderboard | |
| SciUniverse | 14.6% | #5 of 6 | Official leaderboard | |
| FrontierMath Erdős | 0% | #13 of 14 | Epoch AI | |
| FrontierPhysics | 0% | #19 of 19 | Official leaderboard | |
| BrokenArXiv | 25% | #6 of 6 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| WeirdML | 84.8% | #8 of 112 | Official leaderboard | |
| DeepSWE | 73.8% | #2 of 28 | Official leaderboard | |
| StructEval | 80.6% | #2 of 28 | Official leaderboard | |
| SWE Atlas Test Writing | 53.7% | #4 of 22 | Official leaderboard | |
| WebDev Arena | 1,568 | #27 of 105 | Epoch AI | |
| Code Arena | 1,583 | #32 of 124 | Official leaderboard | |
| Vibe Code Bench | 78.7% | #31 of 115 | Official leaderboard | |
| SWE Atlas | 47% | #6 of 22 | Official leaderboard | |
| ALE-Bench | 1,270 | #36 of 121 | Epoch AI | |
| Image-to-WebDev Arena | 1,593 | #18 of 55 | Official leaderboard | |
| SWE Atlas Refactoring | 44.8% | #7 of 17 | Official leaderboard | |
| IOI v2 (Vals) | 56.9% | #20 of 45 | Official leaderboard | |
| Code Migration | 36.5% | #36 of 78 | Official leaderboard | |
| Next.js Agent Evals | 90% | #7 of 14 | Official leaderboard | |
| Vibe Code Bench 1-100 | 18.8% | #14 of 26 | Official leaderboard | |
| FrontierCode | 41.2% | #25 of 46 | Epoch AI | |
| ScarfBench | 66% | #6 of 10 | Official leaderboard | |
| MirrorCode | 38.9% | #18 of 28 | Epoch AI | |
| AutoResearchExam | 0.48 | #6 of 9 | Official leaderboard | |
| Vals RSI Index | 0.20 | #17 of 25 | Official leaderboard | |
| WeirdML v3 | 7.38% | #18 of 25 | Official leaderboard | |
| SWE-Bench Pro V2 | 94.9% | #12 of 16 | Official leaderboard | |
| Code Migration v1.1 | 38.6% | #22 of 29 | Official leaderboard | |
| FrontierSWE | 19.6% | #16 of 21 | Epoch AI | |
| RSI-Exam | 34.1% | #14 of 18 | Official leaderboard | |
| CursorBench | 39.6% | #15 of 19 | Epoch AI | |
| SWE-Bench Pro V2 Hard | 58.8% | #13 of 16 | Official leaderboard | |
| ViBench 2.0 | 0% | #8 of 9 | Official leaderboard | |
| OpenDesign Arena | 68.9 | #17 of 19 | Official leaderboard | |
| Android Bench 2.0 | 8.00% | #10 of 11 | Official leaderboard | |
| Web Clone Bench | 1.00% | #9 of 9 | Official leaderboard | |
| DeepRSI | 643 | #7 of 7 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Game Arena Chess | 1,453 | #1 of 39 | Official leaderboard | |
| Game Arena Dark Hex | 612 | #1 of 38 | Official leaderboard | |
| LLM Chess | 1,546 | #6 of 124 | Official leaderboard | |
| Game Arena Checkers | 1,233 | #2 of 38 | Official leaderboard | |
| Game Arena Four in a Row | 691 | #2 of 38 | Official leaderboard | |
| Game Arena Reversi | 943 | #2 of 36 | Official leaderboard | |
| Game Arena Coin Game | 488 | #2 of 35 | Official leaderboard | |
| Terminal-Bench 2.1 | 89.4% | #5 of 53 | Model card | |
| Kaggle Game Arena | 429 | #5 of 39 | Official leaderboard | |
| Chess Puzzles | 58% | #12 of 83 | Epoch AI | |
| Game Arena Word Art | 1,389 | #6 of 35 | Official leaderboard | |
| Mystery Game Puzzles | 47% | #15 of 80 | Epoch AI | |
| Game Arena Clobber | 478 | #7 of 36 | Official leaderboard | |
| Game Arena Word Association | 338 | #7 of 36 | Official leaderboard | |
| ARC-AGI-3 | 35% | #5 of 24 | Official leaderboard | |
| Game Arena Dots and Boxes | 1,028 | #8 of 37 | Official leaderboard | |
| Game Arena Go | 1,217 | #8 of 36 | Official leaderboard | |
| Game Arena Nine Men's Morris | 779 | #10 of 36 | Official leaderboard | |
| ApprenticeBench | 24% | #6 of 18 | Official leaderboard | |
| Game Arena Poker | 4.68 | #11 of 31 | Official leaderboard | |
| Agent Arena | 0.03 | #19 of 53 | Official leaderboard | |
| Game Arena Bargaining | 214 | #13 of 36 | Official leaderboard | |
| Game Arena Ultimate Tic-Tac-Toe | 380 | #13 of 36 | Official leaderboard | |
| Game Arena Lines of Action | 549 | #15 of 36 | Official leaderboard | |
| Time Horizon Index: KSP | 18.8% | #7 of 13 | Official leaderboard | |
| OSWorld 2.0 | 59% | #11 of 19 | Model card | |
| WitnessBench | 6.70% | #7 of 12 | Official leaderboard | |
| Long-Horizon-Terminal-Bench 2.0 | 47.4% | #9 of 15 | Official leaderboard | |
| Terminal-Bench 4.0 | 19.1% | #17 of 23 | Official leaderboard | |
| Hermes Bench | 62.2% | #15 of 17 | Official leaderboard | |
| CUA-bench | 4.17% | #9 of 9 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vals Finance Agent v2 | 61.4% | #2 of 80 | Official leaderboard | |
| LegalBench | 87% | #8 of 138 | Official leaderboard | |
| Excel Modeling Benchmark | 72.2% | #8 of 76 | Official leaderboard | |
| FailureSensorIQ | 69% | #3 of 24 | Official leaderboard | |
| Vals Finance Agent v2.1 | 65.8% | #6 of 35 | Official leaderboard | |
| APEX-Agents | 64.3% | #12 of 56 | Epoch AI | |
| Excel Modeling Benchmark v1.1 | 72.9% | #8 of 37 | Official leaderboard | |
| Legora BAR | 103.0 | #4 of 17 | Official leaderboard | |
| Harvey LAB (Legal Agent Benchmark) | 10% | #19 of 80 | Official leaderboard | |
| Tax Agent Bench | 32.1% | #21 of 79 | Official leaderboard | |
| TaxEval v2 | 74.5% | #36 of 129 | Official leaderboard | |
| Vals Finance Index | 60.5% | #10 of 35 | Official leaderboard | |
| RCM-Bench-1 | 90.2% | #6 of 20 | Official leaderboard | |
| OR-License | 82.6% | #3 of 10 | Official leaderboard | |
| TaxCalcBench v2 | 61% | #8 of 26 | Official leaderboard | |
| MedCode | 48.1% | #30 of 96 | Official leaderboard | |
| FrontierFinance | 50% | #5 of 16 | Official leaderboard | |
| PRBench | 48.9% | #15 of 44 | Official leaderboard | |
| MedScribe | 84.5% | #34 of 98 | Official leaderboard | |
| Legal Research Bench | 38.9% | #34 of 88 | Official leaderboard | |
| PolicyBench | 85.6% | #21 of 50 | Official leaderboard | |
| Public Benefits Bench | 65.3% | #22 of 50 | Official leaderboard | |
| Physical-Engineering's Last Exam | 0.28 | #8 of 18 | Official leaderboard | |
| Parametric CAD Bench | 36.2% | #5 of 11 | Official leaderboard | |
| Remote Labor Index | 5.83% | #8 of 17 | Official leaderboard | |
| MortgageTax | 65.3% | #41 of 86 | Official leaderboard | |
| Harvey LAB v1.1 (HLAB-Vals-1.1) | 12.1% | #16 of 33 | Official leaderboard | |
| Vals Legal Index | 69.8% | #16 of 33 | Official leaderboard | |
| EEBench | 55.4% | #13 of 26 | Official leaderboard | |
| GDPval-AA v2 | 1,545 | #13 of 26 | Model card | |
| CAD Arena | 0.41 | #6 of 12 | Official leaderboard | |
| Vending-Bench 2 | $5,094 | #34 of 66 | Official leaderboard | |
| Clipping Bench | $596 | #10 of 19 | Official leaderboard | |
| Quant-Arena | 43.0 | #5 of 9 | Official leaderboard | |
| Vals Index v2.2 | 56% | #18 of 32 | Official leaderboard | |
| Build-A-Chip-Bench | 0.80% | #3 of 5 | Official leaderboard | |
| Ivo-micro1 ContractBench | 55% | #5 of 7 | Official leaderboard | |
| LegalReasonBench | 39.6% | #6 of 8 | Official leaderboard | |
| Immigration Bench | 63.4% | #19 of 24 | Official leaderboard | |
| Thrive Holdings helpdesk ticket resolution benchmark | 36.5% | #12 of 15 | Official leaderboard | |
| BlenderWorld | 34.2 | #12 of 13 | Official leaderboard | |
| RealmPathology | 66.3% | #13 of 13 | Official leaderboard | |
| PharmaBench (micro1) | 5.60% | #12 of 12 | Official leaderboard | |
| MissionReport | 54.3% | #9 of 9 | Official leaderboard | |
| Short Form Bench | $1,940 | #9 of 9 | Official leaderboard | |
| micro1 Medical Index | 60% | #8 of 8 | Official leaderboard | |
| Epoch Automation Reports | 42.3% | #6 of 6 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Audio MC | 60.4% | #1 of 16 | Official leaderboard | |
| Vision Arena | 1,291 | #10 of 136 | Official leaderboard | |
| LVBench | 87.8% | #2 of 25 | Model card | |
| GeoBench | 86% | #4 of 48 | Official leaderboard | |
| 1H-VideoQA | 87.1% | #2 of 20 | Official leaderboard | |
| ArtifactArena | 1,223 | #3 of 21 | Official leaderboard | |
| ClockBench | 66.7% | #6 of 37 | Official leaderboard | |
| ParseBench | 71.5% | #7 of 42 | Official leaderboard | |
| OmniParseBench | 92.9 | #1 of 6 | Official leaderboard | |
| Humanity's Sixth Sense | 41.6% | #6 of 29 | Official leaderboard | |
| CharXiv | 86.2% | #16 of 67 | Model card | |
| Blueprint-Bench 2 | 38.6% | #9 of 35 | Official leaderboard | |
| VoxelBench | 1,799 | #22 of 56 | Official leaderboard | |
| Dual-Arm Sim | 38.8 | #4 of 10 | Official leaderboard | |
| GDP.pdf | 23.4% | #17 of 39 | Epoch AI | |
| RoboDojo-RC Tier 1 | 36.2% | #5 of 9 | Official leaderboard | |
| Embodied First Exam | 6.14% | #4 of 7 | Official leaderboard | |
| Furniture Assembly Benchmark | 31.7% | #25 of 35 | Epoch AI | |
| VibeVideoBench | 2.80 | #8 of 9 | Official leaderboard | |
| Code2Video Benchmark | 1,440 | #18 of 19 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BioSecBench-Refusal | 47.4% | #5 of 22 | Official leaderboard | |
| Human Pathogen Capabilities Test v2 (HPCT v2) | 64.4% | #18 of 71 | Official leaderboard | |
| SRE-Bench | 11.1% | #11 of 33 | Official leaderboard | |
| TxBench-Antibody-Discovery | 49% | #10 of 28 | Official leaderboard | |
| scBench-Long | 19.9% | #12 of 27 | Official leaderboard | |
| BioSecBench-Function | 36% | #14 of 27 | Official leaderboard | |
| SpatialBench-Long | 28.6% | #14 of 27 | Official leaderboard | |
| TxBench-Oligo-Discovery | 35.1% | #15 of 28 | Official leaderboard | |
| TxBench-Preclinical-Pharmacology | 54.7% | #15 of 28 | Official leaderboard | |
| PCT (Polarized Contrastive Pairs) | 46.3% | #26 of 46 | Official leaderboard | |
| DistressBench | 70.7% | #24 of 42 | Official leaderboard | |
| BioSecBench-Surveillance | 44.1% | #16 of 27 | Official leaderboard | |
| EpiBench | 21.4% | #18 of 27 | Official leaderboard | |
| VariantBench | 31.4% | #18 of 27 | Official leaderboard | |
| OncologyBench | 13.3% | #23 of 30 | Official leaderboard | |
| MetagenomicsBench | 26.7% | #22 of 26 | Official leaderboard | |
| CyberBench (Vals) | 43.8% | #23 of 26 | Official leaderboard | |
| CyberBench v1.1 (Vals) | 43.8% | #39 of 43 | Official leaderboard | |
| WellnessBench | 47% | #11 of 12 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| CAIS Vision Capabilities Index | 69.9 | #6 of 53 | Official leaderboard | |
| ACCoRD | 77.5 | #18 of 150 | Official leaderboard | |
| CAIS Text Capabilities Index | 50.2 | #8 of 66 | Official leaderboard | |
| Political Science LLM Benchmark | 68.7% | #4 of 29 | Official leaderboard | |
| LMCA | 52.9% | #20 of 139 | Epoch AI | |
| Conceptual Reasoning Index | 65.0 | #22 of 149 | Official leaderboard | |
| SAGE v1.1 (Vals) | 88.2% | #8 of 28 | Official leaderboard | |
| CAIS Risk Index | 49.3 | #29 of 65 | Official leaderboard | |
| Hermes Index | 38.1 | #9 of 17 | Official leaderboard | |
| Vals Science Index | 35.8% | #14 of 23 | Official leaderboard | |
| LiveBench | 59.2% | #22 of 28 | Official leaderboard | |
| SAGE (Vals) | 35.1% | #66 of 81 | Official leaderboard |
Google DeepMind'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. 5 of them aren't in the launch post.