Image to Video Leaderboard (With Audio)Artificial Analysis
Agnes-Video-V2.0
Range | Creator | Model | Elo | 95% CI | Samples | Released | API Pricing 1 | |
|---|---|---|---|---|---|---|---|---|
| 1 | 1 |  ByteDance Seed | Dreamina Seedance 2.0 720p | 1,192 | -10/10 | 6,789 | Mar 2026 | No API available |
| 2 | 2 |  Alibaba-ATH | HappyHorse-1.0 | 1,169 | -10/10 | 6,643 | Apr 2026 | Coming soon |
| 3 | 3-5 |  Google | Veo 3.1 | 1,091 | -9/9 | 5,672 | Jan 2026 | $24.00 /min |
| 4 | 3-6 |  xAI | grok-imagine-video | 1,091 | -9/9 | 6,092 | Jan 2026 | $4.20 /min |
| 5 | 3-8 | Google | Veo 3.1 Fast | 1,084 | -9/9 | 5,084 | Jan 2026 | $9.00 /min |
| 6 | 4-8 |  KlingAI | Kling 3.0 1080p (Pro) | 1,082 | -9/9 | 5,547 | Feb 2026 | $20.16 /min |
| 7 | 5-8 |  PixVerse | PixVerse V6 | 1,081 | -9/9 | 6,766 | Mar 2026 | $6.90 /min |
| 8 | 5-8 | KlingAI | Kling 3.0 720p (Standard) | 1,075 | -9/9 | 5,650 | Feb 2026 | $15.60 /min |
| 9 | 9-12 |  Vidu | Vidu Q3 Pro | 1,065 | -9/9 | 6,386 | Jan 2026 | $9.60 /min |
| 10 | 9-12 | Google | Veo 3.1 Lite | 1,063 | -10/10 | 4,577 | Mar 2026 | $4.80 /min |
| 11 | 9-12 | KlingAI | Kling 3.0 Omni 720p (Standard) | 1,056 | -9/9 | 5,102 | Feb 2026 | $13.44 /min |
| 12 | 9-12 | KlingAI | Kling 3.0 Omni 1080p (Pro) | 1,056 | -9/9 | 5,071 | Feb 2026 | $16.80 /min |
| 13 | 13 | KlingAI | Kling 2.6 Pro (January) | 1,014 | -9/9 | 5,908 | Jan 2026 | $8.40 /min |
| 14 | 14-13 | ByteDance Seed | Seedance 1.5 pro | 1,000 | 0/0 | 4,595 | Dec 2025 | $11.86 /min |
| 15 | 15-17 |  Lightricks | LTX-2.3 Pro  Open Weights | 959 | -10/10 | 5,129 | Mar 2026 | $3.60 /min |
| 16 | 15-17 | PixVerse | PixVerse V5.6 | 957 | -10/10 | 4,381 | Feb 2026 | Coming soon |
| 17 | 15-17 | Lightricks | LTX-2.3 Fast Open Weights | 955 | -10/10 | 5,072 | Mar 2026 | $2.40 /min |
| 18 | 18 | Lightricks | LTX-2 Fast Open Weights | 944 | -10/10 | 4,532 | Oct 2025 | $2.40 /min |
| 19 | 18-19 |  Sapiens AI | Agnes-Video-V2.0 | 934 | -12/12 | 3,027 | May 2026 | $0.30 /min |
| 20 | 20 | Alibaba | Wan 2.6 | 902 | -10/10 | 4,628 | Dec 2025 | $9.00 /min |
| 21 | 21 | Lightricks | LTX-2 Pro Open Weights | 891 | -10/10 | 4,554 | Oct 2025 | $3.60 /min |
1 API Pricing reflects the cost to generate 1 minute of 1080p video on the model creator's API at the model's default settings
Frequently Asked Questions
Dreamina Seedance 2.0 720p currently leads among Image to Video models with audio output in the Artificial Analysis Image to Video Arena with an Elo score of 1192.
The top Image to Video models with audio by Elo rating are: 1. Dreamina Seedance 2.0 720p (Elo 1192), 2. HappyHorse-1.0 (Elo 1169), 3. Veo 3.1 (Elo 1091), 4. grok-imagine-video (Elo 1091), 5. Veo 3.1 Fast (Elo 1084). Rankings are based on blind user votes in the Artificial Analysis Video Arena.
LTX-2.3 Pro currently leads among open weights Image to Video models with audio in the Artificial Analysis Image to Video Arena with an Elo score of 959, followed by LTX-2.3 Fast (Elo 955) and LTX-2 Fast (Elo 944).
HappyHorse-1.0 currently leads the Artificial Analysis Image to Video Arena (without audio) with an Elo score of 1400.
The top Image to Video models without audio by Elo rating are: 1. HappyHorse-1.0 (Elo 1400), 2. Dreamina Seedance 2.0 720p (Elo 1345), 3. grok-imagine-video (Elo 1329), 4. PixVerse V6 (Elo 1326), 5. Kling 2.5 Turbo 1080p (Elo 1288). Rankings are based on blind user votes in the Artificial Analysis Video Arena.
Cosmos3-Super-Image2Video currently leads among open weights Image to Video models without audio in the Artificial Analysis Image to Video Arena with an Elo score of 1255, followed by LTX-2 Pro (Elo 1194) and LTX-2 Fast (Elo 1182).
Text to Video models generate videos from text descriptions alone, while Image to Video models take an existing image as input and animate or extend it into a video. This allows for more control over the visual content and style of the generated video.
Models are ranked using an Elo rating system derived from user votes in blind comparisons. Users compare videos generated from the same input image and choose the result they prefer. Higher Elo scores indicate a model is preferred more often. Vote in the Artificial Analysis Video Arena