Video-MME leaderboard
As of 6 Nov 2026, the top score on Video-MME is 79.7% by video-SALMONN 2+ (ByteDance), ahead of Gemini 1.5 Pro 001 at 75.0%, across 42 model configurations with a published result (Video-MME, result dated 18 Jun 2025). Questions about videos from seconds to an hour long, answered without subtitles. Video-MME is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 6 Nov 2026)
- video-SALMONN 2+ holds the top Video-MME score at 79.7% (Video-MME, 18 Jun 2025).
- Gemini 1.5 Pro 001 is second on Video-MME at 75.0% (Video-MME, 14 May 2024).
- Video-MME: Video understanding across durations, without subtitles (Video-MME).
- 42 model configurations have a published Video-MME result on BenchLeader (BenchLeader, 6 Nov 2026).
- Video-MME does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
Video understanding across durations, without subtitles.
- Published by
- Video-MMEdata via Epoch AI Benchmarking Hub
- Category
- Multimodal
- Index weight
- Reference only
- Models
- 42
- Data as of
- 6 Nov 2026
CC BY 4.0 — Epoch AI, ‘AI Benchmarking Hub’, epoch.ai/benchmarks. Mirrored boards credit their original publishers.
What the test looks like
Questions about videos from seconds to an hour long, answered without subtitles.
How it is scored
Accuracy, as published by Video-MME.
What to keep in mind
Video-capable models only.
- 1video-SALMONN 2+79.7%
- 2Gemini 1.5 Pro 00175.0%
- 3Gemini 1.5 Pro 001 Feb2475.0%
- 4Qwen2.5 VL 72B Instruct73.5%
- 5InternVL2_5-78B72.1%
- 6GPT-4o71.9%
- 7Qwen2-VL-72B71.2%
- 8LLaVA-Video-72B-Qwen270.6%
- 9Gemini 1.5 Flash 00170.3%
- 10ViLAMP-llava-qwen67.5%
- 11Oryx-1.5-32B67.3%
- 12LLaVA-OneVision 72B66.3%
- 13VideoLLAMA3-7B66.2%
- 14LLaVA-Video-7B-Qwen265.9%
- 15LLaVA-Video-7B-Qwen2-TPO65.6%
42 of 42
| # | ||||
|---|---|---|---|---|
| 1 |  video-SALMONN 2+ | 79.7% | – | 2025-06-18 |
| 2 |  Gemini 1.5 Pro 001 | 75.0% | – | 2024-05-14 |
| 3 | Gemini 1.5 Pro 001 Feb24 | 75.0% | – | 2024-02-15 |
| 4 |  Qwen2.5 VL 72B Instructopen ↗ | 73.5% | – | 2024-09-19 |
| 5 |  InternVL2_5-78B | 72.1% | – | 2024-12-06 |
| 6 |  GPT-4o | 71.9% | 43.3 | 2024-11-20 |
| 7 | Qwen2-VL-72B | 71.2% | – | 2024-08-29 |
| 8 |  LLaVA-Video-72B-Qwen2 | 70.6% | – | 2024-09-02 |
| 9 | Gemini 1.5 Flash 001 | 70.3% | – | 2024-05-23 |
| 10 | ViLAMP-llava-qwen | 67.5% | – | 2025-05-01 |
| 11 | Oryx-1.5-32B | 67.3% | – | 2024-10-22 |
| 12 | LLaVA-OneVision 72B | 66.3% | – | 2024-08-06 |
| 13 | VideoLLAMA3-7B | 66.2% | – | 2024-01-22 |
| 14 | LLaVA-Video-7B-Qwen2 | 65.9% | – | 2024-09-02 |
| 15 | LLaVA-Video-7B-Qwen2-TPO | 65.6% | – | 2025-01-19 |
| 16 | VideoChat-Flash-Qwen2-7B_res448 | 65.3% | – | 2025-01-11 |
| 17 | GPT-4o mini | 64.8% | 35.4 | 2024-07-18 |
| 18 | ByteVideoLLM-14B | 64.6% | – | 2024-10-13 |
| 19 |  NVILA 8B | 64.2% | – | 2024-12-10 |
| 20 | LiveCC-7B-Instruct | 64.1% | – | 2024-04-12 |
| 21 | Qwen2-VL-7B | 63.9% | – | 2024-08-29 |
| 22 | MiniCPM-o-2_6 | 63.9% | – | 2025-01-12 |
| 23 | VideoLLAMA2-7B | 62.4% | – | 2024-06-12 |
| 24 | InternVL2-40B | 61.2% | – | 2024-07-08 |
| 25 | Minicpm V 2.6 | 60.9% | – | 2024-08-03 |
| 26 |  Claude 3.5 Sonnet | 60.0% | 47.1 | 2024-06-20 |
| 27 | GPT-4V | 59.9% | – | 2023-11-06 |
| 28 | mPLUG-Owl3-7B-241101 | 59.3% | – | 2024-11-26 |
| 29 | VITA-1.5 | 56.1% | – | 2024-12-20 |
| 30 | Video-XL-7B | 55.5% | – | 2024-10-17 |
| 31 |  Video-CCAM-7B-v1.2 | 53.2% | – | 2024-09-29 |
| 32 | LongLLaVA Qwen2 7B | 52.9% | – | 2024-08-30 |
| 33 | LongVA-7B | 52.6% | – | 2024-06-13 |
| 34 | Qwen-VLmax | 51.3% | – | 2024-01-18 |
| 35 | SliME-Llama3-8B | 45.3% | – | 2024-06-02 |
| 36 | Chat-Uni-Vi-7B-v1.5 + 100k SG-WV | 43.2% | – | 2024-06-20 |
| 37 | Qwen-VL | 41.1% | – | 2023-08-20 |
| 38 | Chat-UniVi-7B-v1.5 | 40.6% | – | 2024-04-23 |
| 39 | ShareGPT4Video 8B | 39.9% | – | 2024-05-27 |
| 40 | Video-LLaVA-7B | 39.9% | – | 2023-11-17 |
| 41 |  video_chat2_mistral | 39.5% | – | 2023-11-29 |
| 42 | ST-LLM | 37.9% | – | 2024-03-28 |
More multimodal benchmarks · All LLM benchmarks
Related benchmarks
Cite as: BenchLeader, “Video-MME leaderboard”, data as of 6 Nov 2026, https://www.benchleader.com/benchmarks/video_mme.
Video-MME: questions
- What does Video-MME measure?
- Questions about videos from seconds to an hour long, answered without subtitles. Scores are reported in percent of tasks solved; higher is better.
- What is the top score on Video-MME?
- As of 6 Nov 2026, the highest score on Video-MME is 79.7% by video-SALMONN 2+ (result dated 18 Jun 2025), ahead of Gemini 1.5 Pro 001 at 75.0%. Scores come from Video-MME and are re-read daily.
- Which AI model leads Video-MME?
- video-SALMONN 2+ leads Video-MME with 79.7% as of 6 Nov 2026 (result dated 18 Jun 2025), ahead of Gemini 1.5 Pro 001 at 75.0%.
- How many models have Video-MME results?
- 42 model configurations have a Video-MME result on BenchLeader, all taken from Video-MME via Epoch AI Benchmarking Hub.
- Who runs Video-MME and how often is it updated?
- Video-MME is published by Video-MME. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does Video-MME count toward the BenchLeader Index?
- No. Video-MME is shown for reference but left out of the composite index.