Generate AI videos with the power of open-source technology. Text-to-video, image-to-video, video-to-video, and motion animation — all powered by Wan 2.6.
Use Wan for flexible video pipelines where open model access, multi-mode generation, and cost-aware iteration all matter.
Wan 2.6 excels at cinematic scenes with precise camera control. Use subject + action + environment + camera + mood for best results.

A camel caravan crossing vast sand dunes at golden hour, long shadows stretching behind, dust particles caught in warm light, slow aerial tracking shot from left to right, epic documentary style

A figure standing alone on a rooftop in the rain, city lights blurred in the background, raindrops splashing on the ledge, camera slowly pushes in from wide to medium close-up, melancholic mood

POV from inside a train window, countryside scenery passing by at speed, reflections of the passenger visible in the glass, golden afternoon light, gentle camera shake simulating train movement

Wan 2.6 is ideal for artists and creators exploring AI-generated video art. Its open-source nature allows fine-tuning for unique visual styles — from ink painting animations to abstract motion art.
Artists & Designers
Generate eye-catching short-form video content for TikTok, Instagram Reels, and YouTube Shorts. Text-to-video and image-to-video modes let you produce scroll-stopping clips without a camera or editing skills.
Content Creators
Transform product photos into dynamic video showcases with smooth rotation, lifestyle contexts, and professional lighting. Video-to-video mode can restyle existing footage to match brand aesthetics.
E-Commerce Sellers
As a fully open-source model, Wan 2.6 is perfect for academic research, model fine-tuning, and building custom video generation pipelines. Run locally with full control over weights and inference.
Researchers & DevelopersSelect Text-to-Video, Image-to-Video, or Video-to-Video depending on your source material and creative goal.
Describe the scene, movement, camera angle, and style. Wan 2.6 responds well to specific, structured prompts with clear subject-action-environment descriptions.
Choose resolution (up to 1080p), duration (up to 15s), and select between Wan 2.6 standard or Wan 2.2 Turbo for faster generation at lower cost.
Hit Generate and wait for your video. Preview the result, then download in full quality or adjust your prompt and settings for another generation.
See how Wan 2.6 stacks up against the leading AI video generation models in 2026.
| Capability | ||||
|---|---|---|---|---|
| Max Resolution | 1080p | 4K | 1080p | 4K (upscale) |
| Max Duration | 15s | 15s | 10s | 10-20s |
| Text to Video |
| Image to Video |
| Video to Video |
| Open Source |
| Motion Control | Animate Move/Replace |
Text to Image Seedream 3.0›ByteDance's image model available on Happy Horse. |
ElevenLabs's audio model available on Happy Horse.
Suno's audio model available on Happy Horse.
Minimax's audio model available on Happy Horse.
Minimax's audio model available on Happy Horse.