Gemini 3.5 Flash ranks first on Toolathlon.
Google DeepMind May 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 | |
|---|---|---|---|---|
| Wer wird Millionär? LLM Benchmark | 433333.0 | #3 of 76 | Official leaderboard | |
| Global MMLU Lite | 95.4% | #2 of 43 | Official leaderboard | |
| LMArena (Elo) | 1,478 | #27 of 333 | Official leaderboard | |
| SEA-HELM | 76.3 | #7 of 62 | Official leaderboard | |
| SimpleQA Verified | 70.4% | #8 of 56 | Official leaderboard | |
| Extended NYT Connections | 92.6% | #16 of 97 | Official leaderboard | |
| FACTS Parametric | 73.5% | #9 of 45 | Official leaderboard | |
| Longform Creative Writing (EQ-Bench) | 71.8 | #36 of 119 | Official leaderboard | |
| FACTS Multimodal | 41.4% | #13 of 39 | Official leaderboard | |
| FACTS Benchmark Suite | 52.9% | #16 of 40 | Official leaderboard | |
| Judgemark v4 | 68.8 | #16 of 40 | Official leaderboard | |
| MRCR v2 | 77.3% | #11 of 27 | Model card | |
| Humanity's Last Exam | 40.2% | #60 of 147 | Model card | |
| Document Arena | 1,461 | #21 of 37 | Official leaderboard | |
| BullshitBench V2 | 20% | #88 of 122 | Official leaderboard | |
| FACTS Search | 62.6% | #24 of 31 | Official leaderboard | |
| EQ-Bench 4 | 1,077 | #28 of 32 | Official leaderboard | |
| FACTS Search V2 | 64.3% | #19 of 21 | Official leaderboard | |
| BullshitBench | 9.09% | #100 of 107 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GPQA Diamond | 92.8% | #22 of 219 | Epoch AI | |
| SimpleBench | 76.7% | #12 of 100 | Official leaderboard | |
| DTBench | 91.1% | #25 of 178 | Official leaderboard | |
| FrontierMath | 39% | #11 of 70 | Epoch AI | |
| MathArena (Apex, IMO, USAMO) | 32.3% | #9 of 47 | Official leaderboard | |
| HMMT 2026 | 95.5% | #6 of 31 | Official leaderboard | |
| OTIS Mock AIME 2024-2025 | 95.6% | #41 of 202 | Epoch AI | |
| ARC-AGI-2 | 72.1% | #24 of 95 | Official leaderboard | |
| ARC-AGI | 92.5% | #26 of 92 | Official leaderboard | |
| FrontierMath Tier 4 | 14.6% | #16 of 55 | Epoch AI | |
| CritPt | 13.1% | #50 of 160 | Epoch AI | |
| NatureBench | 12.2% | #7 of 18 | Official leaderboard | |
| ProofBench | 29% | #24 of 59 | Official leaderboard | |
| Surface Evolver Bench | 58.1% | #11 of 26 | Epoch AI | |
| FrontierMath Tiers 1-3 v2 | 62.8% | #40 of 89 | Epoch AI | |
| FrontierMath Tier 4 v2 | 26.8% | #40 of 71 | Epoch AI | |
| AIME 2026 | 95% | #18 of 31 | Official leaderboard | |
| BixBench3 | 42.1 | #12 of 20 | Official leaderboard | |
| ResearchClawBench | 17.9 | #15 of 23 | Official leaderboard | |
| ProofBench v1.1 | 31% | #35 of 51 | Official leaderboard | |
| Verifying the Verifiers (Project APE CRED benchmark) | 68.5% | #17 of 23 | Official leaderboard | |
| ForecastBench | 64.1 | #64 of 86 | Official leaderboard | |
| Project Euler (MathArena) | 82% | #3 of 4 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SWE-bench Verified | 79.3% | #3 of 32 | Epoch AI | |
| OpenHands Index | 62.6% | #7 of 33 | Official leaderboard | |
| WeirdML | 62.6% | #28 of 112 | Official leaderboard | |
| WebDev Arena | 1,506 | #44 of 105 | Epoch AI | |
| Code Arena | 1,499 | #57 of 124 | Official leaderboard | |
| AndroidBench | 71.1% | #15 of 32 | Official leaderboard | |
| ALE-Bench | 911 | #61 of 121 | Epoch AI | |
| Vibe Code Bench | 48.7% | #61 of 115 | Official leaderboard | |
| Code Migration | 26.8% | #45 of 78 | Official leaderboard | |
| SWE-Marathon | 7.00% | #7 of 12 | Official leaderboard | |
| SWE-Bench Pro | 55.1% | #48 of 77 | Model card | |
| GBAEval | 6.70% | #14 of 21 | Official leaderboard | |
| Image-to-WebDev Arena | 1,504 | #38 of 55 | Official leaderboard | |
| ProgramBench | 0% | #20 of 25 | Official leaderboard | |
| DeepSWE | 36.1% | #25 of 28 | Official leaderboard | |
| Monthly-SWEBench | 29% | #11 of 12 | Official leaderboard | |
| Senior SWE-Bench | 6.32% | #22 of 23 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Toolathlon | 56.5% | #1 of 50 | Official leaderboard | |
| SnakeBench | 34 | #13 of 183 | Official leaderboard | |
| Game Arena Dark Hex | 438 | #4 of 38 | Official leaderboard | |
| Pencil Puzzle Bench | 41.9% | #6 of 53 | Official leaderboard | |
| Game Arena Chess | 1,343 | #5 of 39 | Official leaderboard | |
| MCP Atlas | 83.6% | #6 of 34 | Official leaderboard | |
| LLM Chess | 953 | #25 of 124 | Official leaderboard | |
| LisanBench | 4123.0 | #28 of 133 | Official leaderboard | |
| Game Arena Four in a Row | 551 | #8 of 38 | Official leaderboard | |
| Game Arena Reversi | 712 | #10 of 36 | Official leaderboard | |
| Game Arena Five in a Row | 786 | #6 of 20 | Official leaderboard | |
| AndroidWorld | 93.1% | #3 of 10 | Official leaderboard | |
| Game Arena Word Association | 290 | #11 of 36 | Official leaderboard | |
| Chess Puzzles | 45% | #28 of 83 | Epoch AI | |
| Mystery Game Puzzles | 32% | #28 of 80 | Epoch AI | |
| OSWorld-Verified | 78.4% | #13 of 37 | Model card | |
| CEO-Bench | $75,126 | #7 of 19 | Official leaderboard | |
| Game Arena Word Art | 642 | #13 of 35 | Official leaderboard | |
| Game Arena Coin Game | 382 | #14 of 35 | Official leaderboard | |
| Toolathlon-Verified | 67.3% | #10 of 25 | Official leaderboard | |
| Kaggle Game Arena | 321 | #18 of 39 | Official leaderboard | |
| PinchBench 2.0 | 74.2% | #27 of 57 | Official leaderboard | |
| Game Arena Dots and Boxes | 904 | #19 of 37 | Official leaderboard | |
| Game Arena Checkers | 920 | #20 of 38 | Official leaderboard | |
| Game Arena Ultimate Tic-Tac-Toe | 263 | #20 of 36 | Official leaderboard | |
| Terminal-Bench 2.1 | 76.2% | #30 of 53 | Model card | |
| Game Arena Go | 758 | #21 of 36 | Official leaderboard | |
| Game Arena Clobber | 209 | #22 of 36 | Official leaderboard | |
| Game Arena Lines of Action | 456 | #22 of 36 | Official leaderboard | |
| SkillsBench | 48.2% | #13 of 21 | Official leaderboard | |
| Game Arena Nine Men's Morris | 520 | #26 of 36 | Official leaderboard | |
| Vals Web Search Index | 41.4% | #3 of 4 | Official leaderboard | |
| GAIA | 24.9% | #43 of 57 | Official leaderboard | |
| Game Arena Bargaining | 147 | #30 of 36 | Official leaderboard | |
| EBR-bench | 4.76% | #26 of 28 | Epoch AI | |
| Game Arena Poker | -43.25 | #30 of 31 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MedCode | 55.8% | #6 of 96 | Official leaderboard | |
| Vals Finance Agent v2 | 57.9% | #13 of 80 | Official leaderboard | |
| ProfBench | 56% | #14 of 72 | Official leaderboard | |
| MortgageTax | 68.1% | #19 of 86 | Official leaderboard | |
| GDPval-AA | 1,656 | #4 of 16 | Model card | |
| TaxEval v2 | 74.4% | #37 of 129 | Official leaderboard | |
| CorpFin v2 | 64.7% | #36 of 119 | Official leaderboard | |
| Excel Modeling Benchmark | 63.5% | #28 of 76 | Official leaderboard | |
| LegalBench | 83.6% | #52 of 138 | Official leaderboard | |
| Vending-Bench 2 | $5,396 | #31 of 66 | Official leaderboard | |
| Legal Research Bench | 30.8% | #50 of 88 | Official leaderboard | |
| Harvey LAB (Legal Agent Benchmark) | 2.50% | #46 of 80 | Official leaderboard | |
| Tax Agent Bench | 21.6% | #48 of 79 | Official leaderboard | |
| Public Benefits Bench | 59.5% | #33 of 50 | Official leaderboard | |
| MedScribe | 76.6% | #68 of 98 | Official leaderboard | |
| E-Commerce Bench | 1.90 | #14 of 19 | Official leaderboard | |
| PolicyBench | 79.8% | #37 of 50 | Official leaderboard | |
| JobBench | 31.5% | #12 of 16 | Official leaderboard | |
| RedlineBench | 45.1% | #3 of 4 | Official leaderboard | |
| EEBench | 34.3% | #21 of 26 | Official leaderboard | |
| $OneMillion-Bench | 17% | #11 of 13 | Official leaderboard | |
| TaxCalcBench v2 | 2.00% | #24 of 26 | Official leaderboard | |
| CommerceAgentBench | 43% | #12 of 13 | Official leaderboard | |
| APEX-Agents | 27.5% | #52 of 56 | Epoch AI | |
| Clipping Bench | $218 | #19 of 19 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MMMU-Pro | 83.6% | #2 of 72 | Model card | |
| GeoBench | 85% | #6 of 48 | Official leaderboard | |
| ZeroBench | 22% | #9 of 66 | Official leaderboard | |
| ParseBench | 73.4% | #6 of 42 | Official leaderboard | |
| Vision Arena | 1,282 | #20 of 136 | Official leaderboard | |
| Math Kangaroo 2025 (MathArena) | 89.9% | #4 of 22 | Official leaderboard | |
| 1H-VideoQA | 80.2% | #5 of 20 | Official leaderboard | |
| CharXiv | 84.2% | #19 of 67 | Model card | |
| Blueprint-Bench 2 | 33.6% | #12 of 35 | Official leaderboard | |
| ClockBench | 34.4% | #15 of 37 | Official leaderboard | |
| Humanity's Sixth Sense | 32.6% | #12 of 29 | Official leaderboard | |
| VoxelBench | 1,640 | #31 of 56 | Official leaderboard | |
| GDP.pdf | 14% | #31 of 39 | Epoch AI | |
| EnigmaEval | 25.4% | #4 of 5 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Human Pathogen Capabilities Test v2 (HPCT v2) | 64.7% | #15 of 71 | Official leaderboard | |
| DrugDiscoveryBench | 50% | #9 of 19 | Official leaderboard | |
| CyberBench (Vals) | 69.3% | #13 of 26 | Official leaderboard | |
| DistressBench | 66.9% | #33 of 42 | Official leaderboard | |
| PCT (Polarized Contrastive Pairs) | 53.4% | #38 of 46 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Conceptual Reasoning Index | 59.8 | #33 of 149 | Official leaderboard | |
| SAGE (Vals) | 49.9% | #18 of 81 | Official leaderboard | |
| CAIS Vision Capabilities Index | 64.6 | #12 of 53 | Official leaderboard | |
| LMCA | 47.1% | #32 of 139 | Epoch AI | |
| CAIS Text Capabilities Index | 39.1 | #20 of 66 | Official leaderboard | |
| ACCoRD | 66.6 | #62 of 150 | Official leaderboard | |
| Vals Multimodal Index | 62.9% | #16 of 33 | Official leaderboard | |
| CAIS Risk Index | 61.1 | #52 of 65 | 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.