Wan 2.2 API
Alibaba's Wan 2.2 — open-weight video model deployed on WaveSpeedAI with first-party animate, motion-control, video-edit, speech-to-video, image-to-video, and LoRA-trained variants.
About the Wan 2.2 API
What Wan 2.2 does, how it fits in the Alibaba model lineup, and why teams reach for it.
Wan 2.2 is a video generation model from Alibaba, available through the WaveSpeedAI REST API. Alibaba's Wan 2.2 — open-weight video model deployed on WaveSpeedAI with first-party animate, motion-control, video-edit, speech-to-video, image-to-video, and LoRA-trained variants.
The Wan 2.2 family on WaveSpeedAI ships 35 REST endpoints covering Lora-Support, Image-To-Video, Video-Extend, Motion-Control, Video-To-Video, Image-To-Image, Digital-Human, Training, Text-To-Image, Text-To-Video workflows. Each variant carries its own pricing, parameter knobs, and example outputs — pick the one that matches your input modality and production constraints, or call several from the same API key to compose multi-step pipelines.
Run Wan 2.2 through the same API key, billing account, and rate-limit envelope you use for the other 1,000+ AI models on WaveSpeedAI. No separate vendor setup, no per-provider SDKs, no per-vendor rate-limit envelopes — one integration covers everything from text-to-image and text-to-video through audio synthesis, 3D generation, upscaling, and editing.
All Wan 2.2 API endpoints
35 endpoints available now on WaveSpeedAI — pick the variant that matches your workflow.
Image To Video Lora
Wan-2.2/image-to-video-lora enables unlimited image-to-video generation from a single image, producing smooth, cinematic motion with clean detail. Supports custom LoRAs for style and character consistency. Ready-to-use REST inference API, best performance, no cold starts, affordable pricing.
Image To Video
Wan 2.2 Image-to-Video turns a single image into smooth, cinematic motion with clean detail—ideal for storyboards, mood shots, and product demos. Ready-to-use REST inference API, best performance, no cold starts, affordable pricing.
Video Extend Lora
Extend clips into unlimited longer videos with WAN 2.2 Spicy, producing smooth animation and supporting custom LoRA weights. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Video Extend
Extend short clips into unlimited, high-quality longer videos with smooth animation using WAN 2.2 Spicy. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Image To Video Lora
Generate AI videos with personalized styles using LoRA. Upload images and apply a trained style model to WAN 2.2 — create unique, stylized videos with consistent visual identity.
Image To Video
WAN 2.2 Spicy converts images into unlimited high-quality videos with smooth animations optimized for scalable content generation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Animate
Wan2.2-Animate unified character animation & replacement model replicating movement and expression; generates 720p videos up to 120s. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Video Edit
Wan 2.2 Video Edit lets you modify videos via text prompts (e.g., change clothing or characters). Powered by Wan 2.2, it supports 480p ($0.20/5s) and 720p ($0.40/5s), up to 120s. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Image To Image
WAN 2.2 (14B) is an image-to-image model for high-resolution photorealistic image editing with exceptional precision and fidelity. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Speech To Video
Wan-2.2-S2V turns images and speech into high-fidelity videos with realistic face and body motion; supports up to 10-minute clips in 480p, from $0.15/5s. Ready-to-use REST API, no coldstarts, affordable pricing.
Text To Image Lora
WAN 2.2 generates super-detailed images from text prompts and supports custom LoRAs for fine-grained style and subject control. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Fun Control
Wan2.2-Fun-Control uses Control Codes and multi-modal inputs to generate preset-controlled videos up to 120s at 720p; released under Apache 2.0 for commercial use. Ready-to-use REST API, no coldstarts, affordable.
Wan 2.2 Image Lora Trainer
Train custom Wan 2.2 character/style LoRA models 10x faster. Style training, character training, object training. From concept to model in minutes, not hours. Upload a ZIP file containing images to start!
Wan 2.2 I2v Lora Trainer
Train custom Wan 2.2 I2V LoRA models 10x faster. Action training, motion training, video efect training. From concept to model in minutes, not hours. Upload a ZIP file containing videos to start!
Text To Image Realism
WAN 2.2 delivers ultra-realistic text-to-image generation, converting prompts into photoreal images with high fidelity and detail. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 5b 720p Lora
Wan 2.2 i2v-5B-720p is a 5B image-to-video model producing 720p videos with LoRA support for style customization. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 5b 720p Lora
Wan 2.2 T2V 5B is a 5B text-to-video model with LoRA support that generates 720p videos from text prompts for easy personalization. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 480p Lora Ultra Fast (Fast)
Wan 2.2 i2v delivers ultra-fast Image-to-Video at 480p with support for custom LoRAs for tailored styles. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 480p Ultra Fast (Fast)
Wan 2.2 A14B Image-to-Video (i2v-480p) produces ultra-fast 480p videos from single images, enabling unlimited AI video generation with high throughput. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 720p Ultra Fast (Fast)
Generate unlimited ultra-fast 720p AI videos from images with Wan 2.2 A14B image-to-video model. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 480p Lora Ultra Fast (Fast)
Ultra-fast Wan 2.2 text-to-video model producing 480p videos with custom LoRA support—generate unlimited AI videos with personalized styles. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 720p Lora Ultra Fast (Fast)
Wan 2.2 i2v 720P is an ultra-fast Image-to-Video model that generates unlimited AI videos and supports custom LoRAs for personalized outputs. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 480p Ultra Fast (Fast)
Wan 2.2 t2v 480p Ultra-Fast generates unlimited AI videos from text prompts at 480p with ultra-fast inference. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 5b 720p
Wan 2.2 T2V 5B is a 720P text-to-video model that generates unlimited AI videos from simple text prompts, producing consistent high-quality 720p outputs. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 5b 720p
Wan 2.2 I2V 5B converts images into high-quality 720P videos using a 5B image-to-video model for AI video generation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 480p
Wan 2.2 A14B converts images into 480p videos, enabling unlimited AI video generation from single images. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 480p Lora
WAN 2.2 A14B Image-to-Video model generates unlimited 480p videos from images and supports custom LoRAs for personalized styles and fine-tuning. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 720p Lora
WAN 2.2 Image-to-Video (i2v) 720p converts images into 720p videos and supports custom LoRAs for style personalization. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
I2v 720p
WAN 2.2 A14B i2v-720p converts images into smooth 720p videos, enabling unlimited AI video generation with the Wan 2.2 image-to-video model. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 480p
Wan 2.2 t2v-480p generates unlimited AI videos from text prompts at 480p resolution, ideal for rapid prototyping and content creation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 480p Lora
WAN 2.2 T2V 480p with LoRA generates text-to-video at 480p and supports custom LoRAs for personalized styles. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 720p
Wan 2.2 t2v-720p converts text prompts into native 720P videos, producing high-quality 720P clips from simple prompts. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 720p Lora
Wan 2.2 T2V 720p with custom LoRA support turns text prompts into 720p AI videos and enables unlimited video generation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 720p Lora Ultra Fast (Fast)
Ultra-fast Wan 2.2 Text-to-Video generates unlimited 720p AI videos with custom LoRAs for personalized styles. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
T2v 720p Ultra Fast (Fast)
WAN 2.2 T2V 720p Ultra-Fast generates high-quality 720p videos from text prompts with unlimited output and ultra-fast throughput. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
See Wan 2.2 in action
Real outputs generated by the Wan 2.2 API. Hover any video to preview, click to open the full-size viewer.
How to use the Wan 2.2 API
Four steps from signup to a finished generation. Full Python, Node.js, and cURL examples are in the API section below.
- 1
Get an API key
Sign up for a WaveSpeedAI account and copy your API key from the dashboard. New accounts come with free starter credits — enough to run the playground a few dozen times before billing kicks in.
- 2
Submit a prediction
POST your input as JSON to https://api.wavespeed.ai/api/v3/wavespeed-ai/wan-2.2/image-to-video. The endpoint returns a prediction id immediately — generations are async so you don't hold an open connection during inference.
- 3
Poll for completion
GET https://api.wavespeed.ai/api/v3/predictions/{request_id}/result every 1-2 seconds. The response includes a status field; keep polling until it flips from "queued" or "processing" to "completed".
- 4
Read the output URL
Once status is "completed", read the URL from data.outputs[0]. The URL points to your generated media on the WaveSpeedAI CDN — image, video, audio, or 3D file depending on the Wan 2.2 variant you called.
What you can build with Wan 2.2
Common workflows developers and creators use the Wan 2.2 API for.
Pose-driven character animation
The Animate variant replays motion from a reference video on any character image — useful for dance, sports, choreography, and character-swap content without mocap rigs.
Targeted video editing
The video-edit endpoint applies prompt-driven edits to existing footage — change weather, swap objects, adjust lighting on a specific clip rather than re-generating from scratch.
Custom-style LoRA training
The LoRA trainer endpoints fine-tune Wan 2.2 on your reference set — produce a custom checkpoint for a specific style, brand look, or character, then call it from the LoRA inference endpoint.
Speech-to-video for talking content
The speech-to-video endpoint generates lip-synced talking footage from an audio file + character reference — alternative to InfiniteTalk for cases where you want the full Wan 2.2 toolkit nearby.
Image-to-video at scale
Standard image-to-video is the most-used variant — generate motion from a single still image. Combine with the Spicy tier when you need LoRA-supported stylization.
Stylized output with LoRA tiers
The wan-2.2-spicy variants load community LoRA checkpoints for stylized outputs — useful when your creative direction needs a specific artistic look the base model doesn't produce.
Tips for prompting Wan 2.2
Practical advice for getting better outputs from Wan 2.2 — drawn from the patterns that work across video models in production pipelines.
Be specific about camera moves
Mention concrete cinematography vocabulary — orbit, dolly-in, push-in, pan-left, crane shot, handheld follow. Generic prompts produce static or arbitrary camera choices; named camera moves map directly to motion intent in the model's training data and dramatically improve shot quality.
Anchor character identity with reference images
If your prompt depends on a specific person, character, or product, upload a reference image alongside the prompt. Without a reference, identity drifts across frames and across shots — the same character ends up looking like a slightly different person each generation.
Describe lighting and time of day
Lighting cues like 'golden hour, soft warm directional light' or 'overcast diffused light, slate-grey sky' improve quality and consistency far more than vague quality modifiers. Lighting is one of the strongest priors the model conditions on.
Use negative prompts to suppress common failure modes
Useful negatives for video: 'frame flicker, motion blur, watermark, text artifacts, distorted hands, low resolution, jpeg compression'. Negative prompts cost nothing and noticeably reduce the rate of generations you'd otherwise re-roll.
Pick the shortest duration that captures your beat
Most prompts work best at 5-8 seconds. Longer clips amplify temporal inconsistencies (subject morphing, environment drift). If you need a 20-second sequence, generate three 6-8 second clips and edit them together — quality stays higher than one long generation.
Match aspect ratio to platform up front
9:16 for TikTok / Reels / Shorts, 16:9 for landscape feeds and YouTube, 1:1 for post grids. Models train slightly differently per aspect ratio — cropping a 16:9 to 9:16 after the fact loses both fidelity and the composition the model intended.
Wan 2.2 API pricing
Pricing is per-output. The final charge scales with the parameters you set in each variant's playground (resolution, duration, output count, references).
Call the Wan 2.2 API
Sign up for an API key at wavespeed.ai/accesskey, then submit a prediction via REST. The playground generates ready-to-paste samples for any combination of inputs.
HTTP example
# 1. Submit a prediction
curl -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/wan-2.2/image-to-video" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $WAVESPEED_API_KEY" \
-d '{}'
# 2. Poll the result until status = "completed"
curl -X GET "https://api.wavespeed.ai/api/v3/predictions/{request_id}/result" \
-H "Authorization: Bearer $WAVESPEED_API_KEY"
# Read the output URL from data.outputs[0].Node.js example
// npm install wavespeed
const WaveSpeed = require('wavespeed');
const client = new WaveSpeed(); // reads WAVESPEED_API_KEY
const result = await client.run("wavespeed-ai/wan-2.2/image-to-video", {});
console.log(result.outputs[0]); // → URL of the generated outputPython example
# pip install wavespeed
import wavespeed
output = wavespeed.run(
"wavespeed-ai/wan-2.2/image-to-video",
{}
)
print(output["outputs"][0]) # → URL of the generated outputWan 2.2 vs alternatives
When to pick Wan 2.2 over similar models on WaveSpeedAI.
Wan 2.2 vs Wan 2.7
Wan 2.7 is the newer architecture with multi-modal input and native 1080p output. Wan 2.2 wins on the breadth of specialized variants exposed on WaveSpeedAI — animate, video-edit, speech-to-video, LoRA training — that don't exist in the 2.7 lineup yet.
Wan 2.2 vs Seedance 2.0
Seedance 2.0 has stronger cinematic motion and native audio synthesis. Wan 2.2 wins on the variant breadth (specialized endpoints) and on having open weights underneath — important for LoRA fine-tuning workflows.
Wan 2.2 vs Kling 3.0 Motion Control
Both ship pose-driven animation. Kling Motion Control has stronger identity preservation; Wan 2.2 Animate has open weights and LoRA support, which matters if you want to fine-tune for a specific character or style.
Wan 2.2 API — Frequently asked questions
Pricing, license, integration — common questions about running Wan 2.2 on WaveSpeedAI.
What is the Wan 2.2 API?
Wan 2.2 is a Alibaba video generation model exposed as a REST API on WaveSpeedAI. Alibaba's Wan 2.2 — open-weight video model deployed on WaveSpeedAI with first-party animate, motion-control, video-edit, speech-to-video, image-to-video, and LoRA-trained variants. You can call it programmatically or try it from the playground linked above.
How do I call the Wan 2.2 API?
Sign up for a WaveSpeedAI account, copy your API key from /accesskey, then POST to https://api.wavespeed.ai/api/v3/wavespeed-ai/wan-2.2/image-to-video with your input as JSON. The endpoint returns a prediction id; poll the prediction endpoint until status flips to "completed", then read the output URL from data.outputs[0]. Full Python / Node.js / cURL examples are above.
How much does the Wan 2.2 API cost?
Wan 2.2 starts at $0.020 per run. The exact cost scales with the parameters you set (resolution, duration, output count, references). The live cost preview next to the Generate button in the playground shows the exact price for your current input.
Which Wan 2.2 variants are available?
WaveSpeedAI hosts 35 Wan 2.2 endpoints: wavespeed-ai/wan-2.2/image-to-video-lora, wavespeed-ai/wan-2.2/image-to-video, wavespeed-ai/wan-2.2-spicy/video-extend-lora, wavespeed-ai/wan-2.2-spicy/video-extend, wavespeed-ai/wan-2.2-spicy/image-to-video-lora, wavespeed-ai/wan-2.2-spicy/image-to-video, wavespeed-ai/wan-2.2/animate, wavespeed-ai/wan-2.2/video-edit, and more. Each variant has its own playground page and pricing.
Can I use Wan 2.2 outputs commercially?
Commercial usage rights follow the Alibaba model license. Most Alibaba models permit commercial output use; see each model's playground page for the specific license summary, and WaveSpeedAI's Terms of Service for platform-level conditions.
Why use Wan 2.2 on WaveSpeedAI instead of going direct?
One API key + one billing account across Wan 2.2 AND 1,000+ other AI models from other providers. No per-vendor SDK setup, no separate rate-limit envelopes, no rewrite-per-vendor integration code. Pricing is typically at parity with or below Alibaba's direct API.
About Alibaba
The team behind Wan 2.2 and the broader Alibaba model lineup on WaveSpeedAI.
Alibaba's Tongyi Lab produces the Wan family of video models and the Qwen family of LLMs. Wan is notable for being released with open weights, full-modality input support (text, image, video, and audio references in a single generation), and consistent strength on physics, motion, and multilingual prompts.
Start building with Wan 2.2 on WaveSpeedAI
Free starter credits on signup. One API key across 1,000+ AI models from Alibaba and every other provider.