wan-2.2-animateWan 2.2 Animate turns a character image and a reference video into a new clip in one of two ways. In animate mode, the default, the character and background from your image are driven by the motion and expressions of the person in the video — dance covers, gestures, a talking head — which is how an illustration, a game character or an avatar gets a performance without motion capture. In replace mode the video keeps its scene, camera and motion and only the person is swapped for your character: a virtual character dropped into live footage, or an existing clip recast. The output is exactly as long as the reference video, at 480p or 720p (default 720p), 24fps, and it keeps the reference video's audio track. Billing is per second of the reference video, rounded up to whole seconds: $0.0742/s at 480p and $0.154/s at 720p — a 5-second reference is $0.371 or $0.77, a 30-second one $2.226 or $4.62 — fixed the moment the task is created, and failed requests are never charged. Measured here: a 3.54-second 1080p reference produced a 3.54-second 832x480 clip in 110 seconds, with the character following the gestures beat for beat. Inputs: the image as a public URL or base64 up to 50MB (one person, facing forward, with a clear face works best); the video as a public https MP4 URL up to 120 seconds and 200MB. We read the video's length before creating the task, so an unreadable, private or expired link returns a 400 immediately instead of failing minutes later. Where it fits: dance and choreography recreation, VTuber and avatar reaction clips triggered from a library of reference moves, and recasting existing footage with a virtual lead. Same endpoint and key as the Wan Spicy image-to-video models — only the model field changes.
One person facing forward with a clear face works best; stylised characters work too. Max 50MB.
MP4, max 100MB here (the API takes up to 200MB by public URL), up to 120s. The output is exactly this long, and so is the bill.
animate: your character and background, moving like the person in the video.
Generated video will appear here
Provide URLs and click Generate
Your character and background, moving and emoting like the person in the reference video — no motion capture
The video keeps its scene, camera and motion; only the person becomes your character
No duration parameter. Billed per second of the reference video, rounded up: $0.0742/s at 480p, $0.154/s at 720p
The reference video's sound stays in the output; the frame follows the reference. Measured: 3.54 s in, 3.54 s out
We read the clip length at creation and fix the price there. Unreadable or private links return a 400 right away; up to 120 seconds per clip
Wan 2.2 Animate is a Video Generation API provided by Alibaba. Wan 2.2 Animate turns a character image and a reference video into a new clip in one of two ways. In animate mode, the default, the character and background from your image are driven by the motion and expressions of the person in the video — dance covers, gestures, a talking head — which is how an illustration, a game character or an avatar gets a performance without motion capture. In replace mode the video keeps its scene, camera and motion and only the person is swapped for your character: a virtual character dropped into live footage, or an existing clip recast. The output is exactly as long as the reference video, at 480p or 720p (default 720p), 24fps, and it keeps the reference video's audio track. Billing is per second of the reference video, rounded up to whole seconds: $0.0742/s at 480p and $0.154/s at 720p — a 5-second reference is $0.371 or $0.77, a 30-second one $2.226 or $4.62 — fixed the moment the task is created, and failed requests are never charged. Measured here: a 3.54-second 1080p reference produced a 3.54-second 832x480 clip in 110 seconds, with the character following the gestures beat for beat. Inputs: the image as a public URL or base64 up to 50MB (one person, facing forward, with a clear face works best); the video as a public https MP4 URL up to 120 seconds and 200MB. We read the video's length before creating the task, so an unreadable, private or expired link returns a 400 immediately instead of failing minutes later. Where it fits: dance and choreography recreation, VTuber and avatar reaction clips triggered from a library of reference moves, and recasting existing footage with a virtual lead. Same endpoint and key as the Wan Spicy image-to-video models — only the model field changes. Through APIMODELS platform, you can access this model via a unified API with transparent pay-as-you-go pricing. Current pricing: 480p: $0.0742, 720p: $0.154.
Quickly generate brand promotion videos for ad campaigns and social media marketing.
Create compelling short-form video content for platforms like TikTok, Instagram, and YouTube.
Generate product feature demonstrations and tutorials to improve user conversion.
Produce course explanations, knowledge explainers, and training videos at low cost.
Wan 2.2 Animate is available through APIMODELS at: 480p: $0.0742, 720p: $0.154. Billing is pay-as-you-go — you only pay for what you generate.
Sign up at APIMODELS, get your API key, and call our unified API endpoint. We provide detailed API documentation with code examples in cURL, Python, and Node.js.
APIMODELS offers the same Wan 2.2 Animate model through our aggregation platform. We provide a unified API interface so you do not need separate accounts for each provider - one API key to access all models.
Per second of the reference video, rounded up to whole seconds: $0.0742/s at 480p and $0.154/s at 720p. A 5-second reference is $0.371 or $0.77, 10 seconds $0.742 or $1.54, 30 seconds $2.226 or $4.62. Both modes cost the same, the charge is fixed when the task is created with nothing added afterwards, and failed requests are not charged. Omitting resolution bills at 720p — pass 480p explicitly to pay less.
**animate** (the default) keeps your image — the character and its background — and drives it with the motion and expressions of the person in the reference video, so the output shows your character dancing, talking or gesturing. **replace** keeps the reference video — scene, camera, motion — and only swaps the person for your character, so the output is the original shot with a new lead. Use animate to give an illustration or avatar a performance; use replace to put a virtual character into live footage or recast an existing clip.
The output is exactly as long as the reference video, to the millisecond; there is no duration parameter, and in our test a 3.54-second reference produced a 3.54-second clip. The video must be a public https MP4 URL of at most 120 seconds and 200MB — no base64, no login-gated or signed-expiry links. We read its length before creating the task to fix the price, so an unreadable link returns a 400 with the reason immediately instead of failing minutes later. The frame follows the reference: a 16:9 reference comes out at 832x480 in 480p.
The output keeps the reference video's audio track (stereo AAC in our test) at 24fps; a silent reference gives a silent clip, the model does not add sound of its own. The character image can be a public URL or base64 up to 50MB. One person, facing forward, with a clear unobstructed face works best, and stylised characters drive fine too — our test animated an elf character from a real person's gestures.
Three uses come up first: **dance and choreography recreation** (one character image follows any dance video); **VTuber and avatar reaction clips** (keep a library of reference moves and generate the reaction a fan triggers); **recasting existing footage** (replace mode puts a virtual character into a live shot). Measured here, a 480p job on a 3.54-second reference took 110 seconds; longer clips and 720p take longer, so design around an async task. Same endpoint and key as the Wan Spicy image-to-video models — only the model field changes.
On APIMODELS, Wan 2.2 Animate runs alongside 60+ models on one API key and one balance, so choosing is about fit, not lock-in. It supports Motion Transfer, Character Replace, Image + Video, 480p / 720p, Keeps Audio, Up to 120s, and you can weigh it on price and capability against other Video Generation models, then switch by changing a single model-name string — no new account or integration. Browse every Video Generation option with live pricing at apimodels.app/models.
Wan 2.2 Animate supports: Motion Transfer, Character Replace, Image + Video, 480p / 720p, Keeps Audio, Up to 120s. See the APIMODELS docs for full parameters and call examples.
Yes. APIMODELS exposes Wan 2.2 Animate through a single unified API and one key — no separate provider accounts, and no need to handle each provider's regional network access yourself.
We support Stripe (Visa, Mastercard, and other international cards) and Alipay. Credits are available instantly after payment.