Video model by Alibaba · Coming soon
Wan 2.2 Speech-to-Video on CraftyPeak
Wan 2.2 Speech-to-Video (S2V) is Alibaba's audio-driven model: give it an image and an audio track, and it animates the character to speak or sing in sync.
Wan 2.2 S2V turns a still image and an audio track into a talking or singing video, with lip and body movement driven by the sound.
It is part of the open Wan 2.2 family.
Premium platform. Cheaper pricing.
What Wan 2.2 Speech-to-Video does on CraftyPeak AI
- Audio-driven: Lip and body movement follow your audio.
- Speech and song: Works for dialogue and singing.
- From one image: Animate a single portrait or character.
- Open weights: Part of the Wan 2.2 family.
Wan 2.2 Speech-to-Video specifications
As announced by Alibaba.
- Inputs: Image plus audio track
- Sound: Uses your audio
- Released: August 2025
Wan 2.2 Speech-to-Video, the CraftyPeak way
- One credit wallet for every model: Switch between video and image models without another subscription. Pay per second or per image, only for what you render.
- Templates you can remix: Start from a public template with its prompt and references ready, then make it yours.
- Characters that stay consistent: Build a character once in Character studio and reuse the same face and look across images and videos.
- Brand Studio for product ads: Turn a product and a brief into UGC-style ads and product shots in one workspace.
- Reels Studio and video editor: Plan reels, then trim, arrange and finish your clips in the CraftyPeak multi-track editor.
- PeakShelf library: Every generation lands in your PeakShelf library, ready to download, remix or reuse as a reference.
CraftyPeak passes provider savings straight to you, so every model is priced to undercut big platforms like Higgsfield and Runway.
Made for
- Talking avatars
- Music performances
- Explainers with a presenter
- Podcast visuals
Prompt tips
- Use a clean, front-facing portrait.
- Supply clear audio with little background noise.
- Keep the framing to head and shoulders for lip sync.
An alternative to Higgsfield, Runway, OpenArt, HeyGen
Looking for a Higgsfield, Runway, OpenArt or HeyGen alternative? CraftyPeak AI gives you one credit wallet for every video and image model, and Wan 2.2 Speech-to-Video joins the line-up as soon as it is available.
Higgsfield alternative · Runway alternative · OpenArt alternative · HeyGen alternative · Krea alternative · Freepik alternative · fal.ai alternative · Pollo AI alternative · Pika alternative · Hailuo alternative · Kling AI alternative · Leonardo AI alternative
Also known as
Wan 2.2 S2V, Wan2.2-S2V-14B, Wan speech to video
Frequently asked questions
Is CraftyPeak AI an alternative to Higgsfield, Runway, OpenArt and HeyGen?
Yes. CraftyPeak AI runs video and image models in one credit wallet with pay-per-use pricing. Wan 2.2 Speech-to-Video is coming; the models linked on this page are available today.
Does Wan 2.2 S2V make its own voice?
No. It animates to the audio you supply.
When will Wan 2.2 Speech-to-Video be on CraftyPeak AI?
It is on the way. This page switches to live prices and a Generate button as soon as Wan 2.2 Speech-to-Video is in the prompt bar.