Live ranking, refreshed every 2 days
Best AI Video Generators
As of September 22, 2026, the top-ranked model on LMArena's text-to-video leaderboard is gemini-omni-1.1-flash by Google with a score of 1,516, 3 points ahead of gemini-omni-flash by Google.
Updated October 14, 2026. Leaderboard data from September 22, 2026.
Ranking: top 10 models
Scores come from blind head-to-head votes on LMArena's text-to-video leaderboard. Rank is LMArena's published rank, which accounts for confidence intervals, so several models can share a rank. The top 10 are shown below, and ranks 11 to 40 follow in the expandable list.
| Rank | Model | Lab | Score | Votes | License |
|---|---|---|---|---|---|
| 1 | gemini-omni-1.1-flash | 1,516 ±15 | 1,784 | Proprietary | |
| 2 | gemini-omni-flash | 1,513 ±9 | 26,576 | Proprietary | |
| 3 | flux-3-video | Bfl | 1,493 ±17 | 1,302 | Proprietary |
| 4 | grok-imagine-video-1.5-agent | xAI | 1,492 ±18 | 1,218 | Proprietary |
| 5 | dreamina-seedance-2.0-720p | ByteDance | 1,479 ±8 | 56,318 | Proprietary |
| 6 | wan3.0 | Alibaba (Qwen) | 1,476 ±13 | 2,598 | Proprietary |
| 7 | dreamina-seedance-2.5-720p | ByteDance | 1,474 ±9 | 8,001 | Proprietary |
| 8 | minimax-h3 | MiniMax | 1,460 ±9 | 11,417 | Open weights |
| 9 | muse-video | Meta | 1,456 ±15 | 2,188 | Proprietary |
| 10 | happyhorse-1.0 | Aorizon | 1,427 ±13 | 22,264 | Proprietary |
Show ranks 11 to 40
Rank Model Lab Score License 11 sora-2-pro OpenAI 1,368 ±7 Proprietary 12 veo-3.1-audio Google 1,364 ±14 Proprietary 13 veo-3.1-audio-1080p Google 1,362 ±10 Proprietary 14 veo-3.1-fast-audio Google 1,362 ±10 Proprietary 15 veo-3.1-fast-audio-1080p Google 1,358 ±10 Proprietary 16 veo-3-fast-audio Google 1,348 ±11 Proprietary 17 sora-2 OpenAI 1,343 ±6 Proprietary 18 grok-imagine-video-720p xAI 1,342 ±6 Proprietary 19 veo-3-audio Google 1,340 ±13 Proprietary 20 wan2.7-t2v Wan 1,337 ±7 Proprietary 21 wan2.6-t2v Alibaba (Qwen) 1,327 ±7 Proprietary 22 seedance-v1.5-pro ByteDance 1,256 ±7 Proprietary 23 veo-3 Google 1,254 ±11 Proprietary 24 veo-3-fast Google 1,248 ±12 Proprietary 25 wan2.5-t2v-preview Alibaba (Qwen) 1,245 ±9 Proprietary 26 pixverse-v5.6 Other 1,240 ±11 Proprietary 27 runway-gen-4.5 Runway 1,225 ±9 Proprietary 28 kling-2.5-turbo-1080p Kling 1,220 ±17 Proprietary 29 kling-2.6-pro Kling 1,216 ±7 Proprietary 30 p-video Other 1,207 ±16 Proprietary 31 ray-3 Luma Ai 1,206 ±22 Proprietary 32 hailuo-2.3 MiniMax 1,206 ±6 Proprietary 33 kling-o1-pro Kling 1,205 ±27 Proprietary 34 hailuo-02-pro MiniMax 1,198 ±13 Proprietary 35 seedance-v1-pro ByteDance 1,191 ±11 Proprietary 36 hailuo-02-standard MiniMax 1,181 ±12 Proprietary 37 kandinsky-5.0-t2v-pro Kandinsky 1,175 ±21 Open weights 38 hunyuan-video-1.5 Tencent 1,169 ±16 Open weights 39 veo-2 Google 1,164 ±16 Proprietary 40 kling-v2.1-master Kling 1,163 ±10 Proprietary
Showing the top 40 of 48 models on this leaderboard. Data from September 22, 2026.
Best model by lab
Each lab's highest-ranked model on this leaderboard, and how far it sits from the top.
| Lab | Best model | Rank | Score | Behind #1 |
|---|---|---|---|---|
| gemini-omni-1.1-flash | 1 | 1,516 | Leader | |
| Bfl | flux-3-video | 3 | 1,493 | 23 pts |
| xAI | grok-imagine-video-1.5-agent | 4 | 1,492 | 24 pts |
| ByteDance | dreamina-seedance-2.0-720p | 5 | 1,479 | 37 pts |
| Alibaba (Qwen) | wan3.0 | 6 | 1,476 | 40 pts |
| MiniMax | minimax-h3 | 8 | 1,460 | 56 pts |
| Meta | muse-video | 9 | 1,456 | 60 pts |
| Aorizon | happyhorse-1.0 | 10 | 1,427 | 89 pts |
Which AI lab leads?
The lab with the most models at the top of the board is the one with the deepest bench, not only the single best model.
ByteDance holds 2 of the top 10 models (20%).
Models in the top 10 by lab
Show data table
| Lab | Top 10 |
|---|---|
| ByteDance | 2 |
| 2 | |
| Alibaba (Qwen) | 1 |
| Aorizon | 1 |
| Bfl | 1 |
| Meta | 1 |
| MiniMax | 1 |
| xAI | 1 |
Open vs proprietary: how big is the gap?
The best open-weight model is minimax-h3 by MiniMax at 1,460, 56 points behind the best proprietary model, gemini-omni-1.1-flash by Google at 1,516. Their 95% confidence intervals do not overlap, so the gap is statistically meaningful.
7 proprietary models rank above the best open-weight model, and 1 of the top 20 is open-weight.
Open-weight here means any license other than Proprietary, including licenses with usage limits. It is a broader definition than the OSI one.
Other leaderboards
LMArena runs separate arenas for different jobs. The best model for one is often not the best for another.
Other arenas
- Text (overall)
- WebDev: claude-opus-5.5-max by Anthropic, 1,813
- Vision: claude-fable-5-high by Anthropic, 1,308
- Search: gpt-5.6-sol-xhigh by OpenAI, 1,256
- Document: claude-fable-5 by Anthropic, 1,499
- Text to image: gpt-image-2.5-sunburst by OpenAI, 1,425
- Image editing: gpt-image-2.5-sunburst by OpenAI, 1,524
Text arena by category
How to read this leaderboard
The text-to-video arena compares clips generated from the same prompt and asks people to pick the better one. These are video models, ranked by the same blind-vote method.
- Scores come from blind votes
- LMArena shows a person two anonymous answers to the same prompt and asks which is better. Those head-to-head votes are fitted with a Bradley-Terry model, the statistical successor to Elo, to give each model one arena score. On the Elo scale LMArena uses, a 100-point lead means the higher model is expected to win about 64% of matchups.
- Confidence intervals decide ties
- Every score carries a 95% confidence interval. When two intervals overlap, the order of those two models is not settled. LMArena's rank column already accounts for this, which is why more than one model can hold the same rank.
- Why it differs from usage share
- A high rank does not mean many people use a model. Usage depends on price, distribution and habit. For adoption rather than quality, see AI market share.
Methodology and source
All rankings are read from the LMArena leaderboard dataset on Hugging Face, published by LMArena (formerly Chatbot Arena) under the Creative Commons Attribution 4.0 license. Credit for the underlying votes, scores and confidence intervals belongs to LMArena. Essentras is not affiliated with LMArena.
This page reads the text_to_video subset, category overall, from the dataset's latest split. Scores, intervals, ranks, vote counts and license labels are shown as published. Essentras adds the grouping by lab and the derived figures on this page: gaps, lab shares, open vs proprietary and, on the main page, the history of who held #1.
Data refreshes automatically every 2 days. If the source cannot be reached, the page falls back to the last saved copy and says so at the top. Models are never merged or renamed; model and lab names appear exactly as LMArena publishes them.
Open-weight means any license other than Proprietary. History is limited to what the dataset records: the style-controlled text leaderboard from August 28, 2024. We do not scrape any other ranking site.
Track how AI models rank and describe your brand with our AI SEO service. Want these models to name your brand? Read our guide to LLM SEO.
Frequently asked questions
What is the best AI model for text-to-video generation?
As of September 22, 2026, gemini-omni-1.1-flash by Google ranks first on LMArena's text-to-video leaderboard with a score of 1,516. Rankings move as models and votes are added, so check the table above for the current order.
Which AI lab leads for text-to-video generation?
ByteDance holds 2 of the top 10 models (20%), ahead of Google (2), Alibaba (Qwen) (1), and Aorizon (1).
Is the best open-source model close to the best proprietary model?
The best open-weight model is minimax-h3 by MiniMax at 1,460, 56 points behind the best proprietary model, gemini-omni-1.1-flash by Google at 1,516. 7 proprietary models rank above the best open-weight model, and 1 of the top 20 is open-weight. Their 95% confidence intervals do not overlap, so the gap is statistically meaningful. "Open" here means any license other than Proprietary, which is broader than the OSI definition.
What does this leaderboard measure?
The text-to-video arena compares clips generated from the same prompt and asks people to pick the better one. These are video models, ranked by the same blind-vote method.
How often does this leaderboard update?
This page re-reads LMArena's dataset every 2 days. LMArena publishes a new snapshot when it has gathered enough votes or added models, so the date at the top shows the latest published data.
Models rank themselves. What do they say about you?
A leaderboard shows which AI models are strongest. It does not show what those models tell your customers about your company. Run the free check and see the answers.