50 models
| Rank | Model | Rating | Rating lower bound | Rating upper bound | Variance | Votes | Category | Source updated |
|---|---|---|---|---|---|---|---|---|
| #1 | Gemini Omni 1.2 Flashgoogle · Proprietary | 1,524.46 | 1,555.7 | 63.5 | 1,714 | overall | 2026-11-09 | |
| #2 | Gemini Omni 1.1 Flashgoogle · Proprietary | 1,497.64 | 1,528.29 | 61.13 | 1,792 | overall | 2026-11-09 | |
| #3 | gemini-omni-flashgoogle · Proprietary | 1,500.98 | 1,517.85 | 18.51 | 46,707 | overall | 2026-11-09 | |
| #4 | FLUX 3 Videobfl · Proprietary | 1,474.02 | 1,508.09 | 75.53 | 1,314 | overall | 2026-11-09 | |
| #5 | Grok Imagine Video 1.5xai · Proprietary | 1,468.91 | 1,506.09 | 89.97 | 1,249 | overall | 2026-11-09 | |
| #6 | Dreamina Seedance 2.0 (720p)bytedance · Proprietary | 1,475.51 | 1,491.26 | 16.15 | 72,137 | overall | 2026-11-09 | |
| #7 | Dreamina Seedance 2.5 (720p)bytedance · Proprietary | 1,471.21 | 1,487.78 | 17.86 | 25,421 | overall | 2026-11-09 | |
| #8 | Wan 3.0alibaba · Proprietary | 1,464.65 | 1,487.76 | 34.76 | 3,060 | overall | 2026-11-09 | |
| #9 | SkyReels V4skyreel · Proprietary | 1,440.43 | 1,477.04 | 87.21 | 1,060 | overall | 2026-11-09 | |
| #10 | Muse Videometa · Proprietary | 1,442.57 | 1,471.69 | 55.21 | 2,225 | overall | 2026-11-09 | |
| #11 | MiniMax H3minimax · minimax-h3-community-license-agreement | 1,447.78 | 1,464.07 | 17.28 | 29,257 | overall | 2026-11-09 | |
| #12 | HappyHorse 1.0aorizon · Proprietary | 1,416.18 | 1,441.14 | 40.56 | 22,896 | overall | 2026-11-09 | |
| #13 | Sora 2 Proopenai · Proprietary | 1,360.34 | 1,373.57 | 11.39 | 55,965 | overall | 2026-11-09 | |
| #14 | Veo 3.1 (Audio)google · Proprietary | 1,348.97 | 1,377.77 | 53.98 | 13,790 | overall | 2026-11-09 | |
| #15 | Veo 3.1 (Audio, 1080p)google · Proprietary | 1,352.38 | 1,372.76 | 27.04 | 25,636 | overall | 2026-11-09 | |
| #16 | Veo 3.1 Fast (Audio)google · Proprietary | 1,350.69 | 1,371.54 | 28.31 | 40,278 | overall | 2026-11-09 | |
| #17 | Veo 3.1 Fast (Audio, 1080p)google · Proprietary | 1,348.97 | 1,368.62 | 25.14 | 26,483 | overall | 2026-11-09 | |
| #18 | Veo 3 Fast (Audio)google · Proprietary | 1,335.93 | 1,358.51 | 33.18 | 25,277 | overall | 2026-11-09 | |
| #19 | Sora 2openai · Proprietary | 1,336.19 | 1,348.29 | 9.53 | 73,992 | overall | 2026-11-09 | |
| #20 | Grok Imagine Video (720p)xai · Proprietary | 1,334.71 | 1,347.17 | 10.1 | 190,109 | overall | 2026-11-09 | |
| #21 | Veo 3 (Audio)google · Proprietary | 1,327.72 | 1,352.94 | 41.4 | 19,032 | overall | 2026-11-09 | |
| #22 | Wan 2.7alibaba · Proprietary | 1,329.94 | 1,343.93 | 12.74 | 48,517 | overall | 2026-11-09 | |
| #23 | Wan 2.6alibaba · Proprietary | 1,316.45 | 1,330.56 | 12.96 | 75,753 | overall | 2026-11-09 | |
| #24 | Seedance 1.5 Probytedance · Proprietary | 1,249.45 | 1,262.81 | 11.62 | 77,992 | overall | 2026-11-09 | |
| #25 | Veo 3google · Proprietary | 1,243.02 | 1,265.5 | 32.89 | 15,017 | overall | 2026-11-09 | |
| #26 | Veo 3 Fastgoogle · Proprietary | 1,238.74 | 1,262.23 | 35.91 | 15,308 | overall | 2026-11-09 | |
| #27 | Wan 2.5 Previewalibaba · Proprietary | 1,235.7 | 1,252.9 | 19.25 | 43,113 | overall | 2026-11-09 | |
| #28 | PixVerse V5.6Proprietary | 1,227.18 | 1,248.1 | 28.48 | 33,302 | overall | 2026-11-09 | |
| #29 | Runway Gen-4.5runway · Proprietary | 1,215.01 | 1,231.26 | 17.2 | 65,376 | overall | 2026-11-09 | |
| #30 | Kling 2.5 Turbo (1080p)kling · Proprietary | 1,202.03 | 1,236.94 | 79.32 | 2,110 | overall | 2026-11-09 | |
| #31 | Kling 2.6 Prokling · Proprietary | 1,209.34 | 1,222.62 | 11.48 | 75,018 | overall | 2026-11-09 | |
| #32 | P-VideoProprietary | 1,190.88 | 1,222.12 | 63.53 | 7,318 | overall | 2026-11-09 | |
| #33 | Luma Ray3luma-ai · Proprietary | 1,183.87 | 1,228.56 | 129.96 | 1,136 | overall | 2026-11-09 | |
| #34 | Hailuo 2.3minimax · Proprietary | 1,199.52 | 1,211.35 | 9.11 | 91,747 | overall | 2026-11-09 | |
| #35 | Kling O1 Prokling · Proprietary | 1,174.23 | 1,227.95 | 187.83 | 1,210 | overall | 2026-11-09 | |
| #36 | Hailuo 02 Prominimax · Proprietary | 1,186.04 | 1,211.17 | 41.1 | 9,399 | overall | 2026-11-09 | |
| #37 | Seedance 1.0 Probytedance · Proprietary | 1,180.14 | 1,203.15 | 34.46 | 12,167 | overall | 2026-11-09 | |
| #38 | Hailuo 02 Standardminimax · Proprietary | 1,169.12 | 1,193.35 | 38.2 | 9,379 | overall | 2026-11-09 | |
| #39 | Kandinsky 5.0 Video Prokandinsky · MIT | 1,154.82 | 1,196.12 | 110.96 | 2,020 | overall | 2026-11-09 | |
| #40 | HunyuanVideo 1.5tencent · tencent-hunyuan-community | 1,152.17 | 1,184.73 | 69.02 | 4,322 | overall | 2026-11-09 | |
| #41 | Veo 2google · Proprietary | 1,148.21 | 1,181 | 69.93 | 6,531 | overall | 2026-11-09 | |
| #42 | Kling 2.1 Masterkling · Proprietary | 1,153.06 | 1,172.93 | 25.7 | 14,104 | overall | 2026-11-09 | |
| #43 | LTX-2 19Bltx-2-community-license-agreement | 1,148.8 | 1,163.44 | 13.94 | 92,462 | overall | 2026-11-09 | |
| #44 | Wan 2.2 A14Balibaba · Apache 2.0 | 1,116.59 | 1,146.98 | 60.11 | 10,458 | overall | 2026-11-09 | |
| #45 | Seedance 1.0 Litebytedance · Proprietary | 1,103.03 | 1,122.31 | 24.2 | 16,258 | overall | 2026-11-09 | |
| #46 | Kandinsky 5.0 Video Litekandinsky · MIT | 1,094.58 | 1,130.43 | 83.65 | 1,472 | overall | 2026-11-09 | |
| #47 | Soraopenai · Proprietary | 1,052.93 | 1,085.61 | 69.5 | 4,082 | overall | 2026-11-09 | |
| #48 | Luma Ray2luma-ai · Proprietary | 1,048.29 | 1,083.15 | 79.09 | 5,250 | overall | 2026-11-09 | |
| #49 | Pika 2.2pika · Proprietary | 993.89 | 1,024.71 | 61.82 | 5,763 | overall | 2026-11-09 | |
| #50 | Mochi 1genmo · Apache 2.0 | 990.26 | 1,023.79 | 73.16 | 5,893 | overall | 2026-11-09 |
Evaluates coding agents on 113 long-horizon, real-world software engineering tasks written from scratch. Compares Pass@1 accuracy against actual cost and step counts while avoiding training contamination.
Opens in a new tabIndependent benchmarks rather than provider-reported results. Its Intelligence Index covers reasoning, coding, science and knowledge alongside speed, pricing and context window metrics.
Opens in a new tabNew questions every month help prevent contamination. Automated, objective grading with unique correct answers focuses on mathematics, coding, reasoning and data analysis; widely recognized in academia.
Opens in a new tabOriginal Chinese questions tailored to Chinese language, policy and culture, covering six capabilities including mathematical reasoning, coding, hallucination control and agent planning.
Opens in a new tabCreated by Shanghai AI Laboratory with over 20 companies, spanning five capability dimensions and 70+ datasets for Chinese and English models, plus human voting in Compass Arena.