Wan 2.2 weights: Apache 2.0, 5-second 480p/720p; Sume hosts wan-3.0
Wan 2.2 T2V-A14B is Apache 2.0 with open weights and makes 5-second 480P or 720P clips. Sume does not host Wan 2.2; its Wan row is wan-3.0, 2 to 30 seconds.

Wan 2.2 is the Wan generation with open weights under Apache 2.0. Its Hugging Face card for T2V-A14B says the model makes 5-second videos at 480P and 720P with a Mixture-of-Experts design, and that code and weights were released on Jul 28, 2025. Sume does not host Wan 2.2. Its Wan row is wan-3.0, an API model with a 2 to 30 second range and 1080p.
Wan 2.2 facts are from the Wan2.2-T2V-A14B card, read 2026-10-02. Sume's are from the OpenAPI schema, Video Router and Video generation. The card says nothing about Wan 3.0.
What does the Wan 2.2 card say you get?
The card lists three checkpoints: T2V-A14B (text-to-video, 480P and 720P), I2V-A14B (image-to-video, 480P and 720P) and TI2V-5B (text and image to video at 720P, 24 fps, built on a high-compression Wan2.2-VAE). The license tag on the page is apache-2.0.
Hardware is not trivial. The card's single-GPU T2V command says it can run on a GPU with at least 80GB of VRAM, and it points to offload and dtype options if you hit out-of-memory errors. The card says the 5B model can run on consumer cards such as a 4090.
| Item | Wan 2.2 (open weights) | Sume wan-3.0 (hosted) |
|---|---|---|
| License / access | apache-2.0 on the card | API access with a Sume key |
| Length | 5 seconds on T2V-A14B | duration 2 to 30 seconds |
| Resolution | 480P and 720P (A14B); 720P (5B) | 480p, 720p, 1080p |
| Hardware | At least 80GB VRAM for the single-GPU T2V command | None on your side |
Is the Wan 2.2 license a problem for commercial use?
The card's tag says Apache 2.0 and links a license agreement section. Read that text yourself before you ship; this post only reports what the page says and is not legal advice. The card also asks that you share projects built on Wan2.1 or Wan2.2 if you want them highlighted, which is a request, not a condition.
When do I call wan-3.0 instead of running Wan 2.2?
Call the hosted model when you need more than 5 seconds, 1080p, or no GPU to run. Send model: "wan-3.0" with a duration from 2 to 30. Run Wan 2.2 yourself when you need the weights, offline use, or to modify the model. The two are different generations, so do not expect the same output from the same prompt.
curl -X POST https://api.sume.com/v1/video-router/generate \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: wan3-001" \
-d '{
"model": "wan-3.0",
"prompt": "A paper boat drifts down a rain gutter, close camera",
"resolution": "720p",
"duration": 12,
"mode": "async"
}'Sources
Related posts
More in Models
- Wan 3.0 allows 20,000 prompt characters, MiniMax H3 7,000
Alibaba allows 20,000 characters in a Wan 3.0 prompt; MiniMax caps each H3 text item at 7,000. How to write one prompt that fits both on Sume.
- Wan 3.0 input video plus output capped at 30 seconds: Sume plan
Alibaba caps Wan 3.0 input video plus output at 30 seconds. What that means for reference_video_urls and duration when you call wan-3.0 on Sume.
- An OpenRouter-compatible video API: sume/auto or a pinned model
Sume's POST /v1/videos follows OpenRouter's video generation API field for field. Let sume/auto pick the model, or pin a catalog id like seedance-2.5.
- Image generation API with reference images: POST /v1/images
Send a prompt plus public HTTPS reference images to Sume's POST /v1/images. Pin a catalog model or send sume/auto; the catalog lists each model's limits.
Written by Sume