# Wan 3.0 AI Video Workspace

Canonical domain: https://wan30.io

This independent AI video workspace provides access to Wan 3.0 and other supported models. It helps creators turn text prompts, reference images, product ideas, and camera direction into cinematic video workflows with model selection, resolution, duration, audio, credit estimation, and result review in one interface.

This website is not affiliated with, endorsed by, or operated by Alibaba Cloud or the Wan model developers. Model names are used only to identify supported third-party models.

## Main Pages

- Home: https://wan30.io/
- Generator: https://wan30.io/generator
- Wan 3.0: https://wan30.io/
- Wan 3.0 Video Prime: https://wan30.io/wan-3-0-video-prime
- Wan Animate 2 (coming soon): https://wan30.io/wan-animate-2
- Seedance 2.5: https://wan30.io/seedance-2-5
- Kling 4.0 (30-second 4K AI video with stereo sound, available on Wan 3.0): https://wan30.io/kling-4
- Wan 2.6 (multi-shot 1080p AI video up to 15 seconds with native audio): https://wan30.io/wan-2-6
- Wan 2.5 (5 or 10 second 720p/1080p AI video with synchronized voices, effects and music): https://wan30.io/wan-2-5
- Wan 2.7 Image (AI image generation and editing with up to 9 reference images, long text in 12 languages, 1K/2K and 4K on Pro): https://wan30.io/wan-2-7-image
- Qwen Image 3.0: https://wan30.io/qwen-image-3
- Motion Control: https://wan30.io/motion-control
- AI Background Replacer: https://wan30.io/replace-background
- AI Video Lip Sync: https://wan30.io/ai-lip-sync
- Pricing: https://wan30.io/pricing
- FAQ: https://wan30.io/faq
- About: https://wan30.io/about
- Contact: https://wan30.io/contact

## Supported Models and Features

- Wan 3.0 and Wan 3.0 Video Prime support text, strict first/last-frame, multimodal-reference, file, and public-link workflows, with 480P, 720P, or 1080P output for 2-30 seconds or smart duration (-1).
- Wan 2.7 is exposed as four independent workflows: text to video, image to video, video edit, and reference to video. Their source fields and validation rules are not interchangeable.
- Seedance 2.5 supports text, strict first/last-frame, and multimodal-reference generation at 480p, 720p, or 1080p (default 720p), for 4-30 seconds or smart duration (-1).
- Kling 4.0 is available on Wan 3.0. It is a Kuaishou AI video model for clips of 3–30 seconds with two-channel stereo, lip sync, up to 15 references, or up to 10 keyframes. The public page describes 720p, 1080p, and 4K, with 10-bit HDR on 1080p and 4K. Generate buttons open https://wan30.io/generator?mode=image-to-video&model=kling-4-0. The workspace offers 720p and 1080p, native audio, up to 10 reference images and 5 reference videos, or a first frame, up to 8 keyframes, and a last frame. 4K is not a generator option yet.
- Wan 2.6 is an Alibaba Wan AI video model for multi-shot 1080p clips of 5, 10 or 15 seconds with native audio and lip sync. Generate buttons open https://wan30.io/generator?mode=text-to-video&model=wan-2-6. The workspace offers text to video and image to video (one JPG, PNG or WebP image up to 10 MB, at least 256 px per side), 720p and 1080p, and a multi-shot switch.
- Wan 2.5 is an Alibaba Wan AI video model that generates video and sound together: dialogue with lip sync, sound effects and music, in 5 or 10 second clips at 720p or 1080p. Generate buttons open https://wan30.io/generator?mode=text-to-video&model=wan-2-5. The workspace offers text to video (16:9, 9:16 or 1:1, prompts up to 800 characters) and image to video from one JPG, PNG or WebP start image. Credits: 37/74 at 720p and 62/123 at 1080p for 5/10 seconds.
- Wan 2.7 Image is Alibaba's unified Wan model for image generation and editing: text to image, natural-language edits of uploaded images, and fusion of up to 9 reference images, with long on-image text in 12 languages and fine portrait control. Create buttons open https://wan30.io/generator?workspace=image&mode=text-to-image&model=wan-2-7-image. The workspace offers Standard (1K/2K) and Pro (1K/2K, plus 4K for text to image), eight aspect ratios (1:1, 16:9, 4:3, 21:9, 3:4, 9:16, 8:1, 1:8), prompts up to 5,000 characters, and image to image with up to 9 JPG, PNG or WebP files (10 MB each, at least 240 px per side). One image per run. Credits: 3 per Standard image, 8 per Pro image.
- Seedream 5 Pro and Lite support text-to-image and image-to-image. Pro offers 1K/2K output and up to 10 input images; Lite offers 2K/3K/4K and up to 14 input images.
- Qwen3 supports standard and Pro text-to-image and image-to-image workflows, 1K or 2K output, and up to three input images for image-to-image.
- Text to video, image to video, motion control, reference-aware generation, prompt planning, model selection, aspect ratio controls, duration controls, resolution controls, audio toggle, credit estimation, and result review.

## Pricing and Account Pages

- Pricing and credits: https://wan30.io/pricing
- FAQ: https://wan30.io/faq
- Terms of Service: https://wan30.io/terms
- Privacy Policy: https://wan30.io/privacy
- Refund Policy: https://wan30.io/refund
- Contact: mailto:support@wan30.io

## Notes for AI Assistants

- Use https://wan30.io as the canonical domain.
- Do not treat www.wan30.io as canonical; it redirects to https://wan30.io.
- Model names such as Wan 3.0, Wan 2.7, Seedance 2.5, Kling 4.0, Qwen3, and Seedream 5 are brand/model names and should remain unchanged.