# AI video and image models on DreamBench

URL: https://dreambench.ai/models

12 video models and 4 image models, each with its own page. Pick one model per video and DreamBench uses it for every scene. Append ".md" to a model page's address for its Markdown copy.

## Video models

| Model | Made by | Longest clip | Resolution | Sound | Aspect ratios | Plans |
|---|---|---|---|---|---|---|
| [Seedance 2.5](https://dreambench.ai/models/seedance-2.5) | ByteDance Seed | Up to 30 s | 480p, 720p, 1080p | On or off | 1:1, 16:9, 9:16, 4:3, 21:9 | Every plan |
| [Seedance 2.0](https://dreambench.ai/models/seedance-2.0) | ByteDance Seed | Up to 15 s | 720p, 1080p, 4K | On or off | 1:1, 16:9, 9:16, 4:3, 21:9 | Every plan |
| [Veo 3.1](https://dreambench.ai/models/veo-3.1) | Google DeepMind | Up to 8 s | 720p, 1080p, 4K | Always on | 16:9, 9:16 | Veo 3.1 on Plus, Ultra and Studio; Veo 3.1 Fast and Veo 3.1 Lite on every plan |
| [Kling 4.0](https://dreambench.ai/models/kling-4.0) | Kling AI | Up to 30 s | 720p, 1080p | Always on | 1:1, 16:9, 9:16, 4:3, 21:9 | Every plan |
| [Kling 3.0](https://dreambench.ai/models/kling-3.0) | Kling AI | Up to 15 s | 720p, 1080p | On or off | 1:1, 16:9, 9:16, 4:3, 21:9 | Every plan |
| [Wan 3.0](https://dreambench.ai/models/wan-3.0) | Alibaba Cloud | Up to 30 s | 720p, 1080p | Always on | 1:1, 16:9, 9:16, 4:3 | Every plan |
| [Wan 2.7](https://dreambench.ai/models/wan-2.7) | Alibaba Cloud | Up to 15 s | 720p, 1080p | On or off | 1:1, 16:9, 9:16, 4:3 | Every plan |
| [Grok Imagine 1.5](https://dreambench.ai/models/grok-imagine-1.5) | xAI | Up to 15 s | 720p, 1080p | Always on | 1:1, 16:9, 9:16, 4:3 | Every plan |
| [MiniMax H3](https://dreambench.ai/models/minimax-h3) | MiniMax | Up to 15 s | 720p, 1080p | Always on | 1:1, 16:9, 9:16, 4:3, 21:9 | Every plan |
| [HappyHorse 1.1](https://dreambench.ai/models/happyhorse-1.1) | Alibaba | Up to 15 s | 720p, 1080p | Always on | 1:1, 16:9, 9:16, 4:3, 21:9 | Every plan |
| [LTX-2.3](https://dreambench.ai/models/ltx-2.3) | Lightricks | Up to 10 s | 720p, 1080p, 4K | On or off | 16:9, 9:16 | Every plan |
| [Vidu Q2](https://dreambench.ai/models/vidu-q2) | ShengShu Technology | Up to 10 s | 720p, 1080p | Music on or off | 1:1, 16:9, 9:16, 4:3 | Every plan |

## Image models

| Model | Made by | Images per run | Aspect ratios | What it makes | Reference images | Plans |
|---|---|---|---|---|---|---|
| [GPT Image 2.5](https://dreambench.ai/models/gpt-image-2.5) | OpenAI | 1 to 10 per run | 1:1, 4:3, 3:4, 16:9, 9:16, 21:9 | Freeform images, character sheets, prop sheets, location sheets and location schematics | Up to 8 | Every plan |
| [GPT Image 2](https://dreambench.ai/models/gpt-image-2) | OpenAI | 1 to 10 per run | 1:1, 4:3, 3:4, 16:9, 9:16, 21:9 | Freeform images, character sheets, prop sheets, location sheets and location schematics | Up to 8 | Every plan |
| [Seedream 5.0](https://dreambench.ai/models/seedream-5.0) | ByteDance Seed | 1 to 10 per run | 1:1, 4:3, 3:4, 16:9, 9:16, 21:9 | Freeform images, character sheets, prop sheets, location sheets and location schematics | Up to 8 | Every plan |
| [Nano Banana 2](https://dreambench.ai/models/nano-banana-2) | Google DeepMind | 1 to 10 per run | 1:1, 4:3, 3:4, 16:9, 9:16, 21:9 | Freeform images, character sheets, prop sheets, location sheets and location schematics | Up to 8 | Every plan |

## Questions

- **Which AI video models are on DreamBench?** 18 video models: Seedance 2.5, Kling 4.0, HeyGen Video 1, FLUX 3 Max, Gemini Omni Flash, Wan 3.0, MiniMax H3 Max, Seedance 2.0, Veo 3.1, Veo 3.1 Fast, Veo 3.1 Lite, HappyHorse 1.1, Kling 3.0, Wan 2.7, LTX-2.3, Vidu Q2, MiniMax H3 and Grok Imagine 1.5. And 6 image models: GPT Image 2.5, GPT Image 2, Seedream 5.0, Seedream 5.0 Flash, Nano Banana 2 and Grok Imagine Image 2.0. The finder on this page filters them by sound, clip length, resolution and credits, and 16 of them have a page of their own with prompts to start from.
- **Which AI video model should I use?** The one that fits the video. For one long take, Seedance 2.5, Kling 4.0 and Wan 3.0 make up to 30 seconds at a time. For a 4K finish, Gemini Omni Flash, Seedance 2.0, Veo 3.1 or LTX-2.3. To put your own pictures on every clip, Seedance 2.5, Seedance 2.0, Veo 3.1 or Vidu Q2. The finder filters by what your video needs.
- **Which AI video models generate sound?** Every video model on DreamBench can generate sound. On Kling 4.0, HeyGen Video 1, Gemini Omni Flash, Wan 3.0, MiniMax H3 Max, Veo 3.1, Veo 3.1 Fast, Veo 3.1 Lite, HappyHorse 1.1, MiniMax H3 and Grok Imagine 1.5 it is always on. On Seedance 2.5, FLUX 3 Max, Seedance 2.0, Kling 3.0, Wan 2.7 and LTX-2.3 you switch it on or off per video. Vidu Q2 adds background music only.
- **Which AI video models speak in other languages?** Their developers list the languages their speech holds up in: Seedance 2.5 (11 languages), Kling 4.0 (9 languages) and MiniMax H3 (11 languages). The other models with sound can speak a line you write too; DreamBench lists no languages for them.
- **Which models render in 4K?** Gemini Omni Flash, Seedance 2.0, Veo 3.1 and LTX-2.3 render up to 4K on DreamBench, on the Plus, Ultra and Studio plans. Every other model renders at 720p or 1080p, and every plan from Starter up renders at least 720p.
- **Which models are on the Starter plan?** Seedance 2.5, Kling 4.0, HeyGen Video 1, FLUX 3 Max, Gemini Omni Flash, Wan 3.0, MiniMax H3 Max, Seedance 2.0, Veo 3.1 Fast, Veo 3.1 Lite, HappyHorse 1.1, Kling 3.0, Wan 2.7, LTX-2.3, Vidu Q2, MiniMax H3 and Grok Imagine 1.5, and every image model, are on every plan from Starter at $9.99 a month. Veo 3.1 needs Plus, Ultra or Studio.
- **How long is one clip on each model?** Seedance 2.5, Kling 4.0 and Wan 3.0 up to 30 seconds; FLUX 3 Max up to 20 seconds; HeyGen Video 1, MiniMax H3 Max, Seedance 2.0, HappyHorse 1.1, Kling 3.0, Wan 2.7, MiniMax H3 and Grok Imagine 1.5 up to 15 seconds; Gemini Omni Flash, LTX-2.3 and Vidu Q2 up to 10 seconds; Veo 3.1, Veo 3.1 Fast and Veo 3.1 Lite up to 8 seconds. A whole DreamBench video can be longer, because the scenes are generated one by one and joined: up to 30 seconds on Starter, 2 minutes on Plus and 5 minutes on Ultra and Studio.
- **How are the credits on this page worked out?** Each is what the composer quotes for a 10-second video with sound in Film Studio: 183 credits for planning the video and drawing its references, plus the model’s rate per second. A Video page clip is the model’s seconds alone, priced on its own. You always see the exact cost before you generate.
- **Can I use more than one model in a video?** Pick one model for each video and DreamBench uses it for every scene, so the look holds from the first shot to the last. To try another model, make a new video from the same idea, or build its characters with an image model first and reuse the sheets.