← All models
Alibaba
New
4.0
188 cr
Alibabaimage → videoNEW

Wan 2.5

Permissive · composed reference

Alibaba Wan 2.5 — the permissive video backbone Higgsfield uses. With an uploaded photo we compose a scene first, then animate. Native audio, 1080p, 5s or 10s.

Price
188 cr
Avg latency
~120s
Tier
premium
fal.ai model id
fal-ai/wan-25-preview/image-to-video

What Wan 2.5 is built for

Wan 2.5 by Alibaba is a premium-tier image to video model used to animate a still image into motion.

Inside Lumineer AI, Wan 2.5 runs on the same unified credit balance as every other model — no per-vendor billing, no API keys to manage. One generation costs 188 credits and completes in roughly 120 seconds.

Teams pick Wan 2.5 for product motion, portrait animation, still-to-reel and similar use cases where Alibaba's aesthetic bar is non-negotiable. Failed generations are refunded automatically, and every output is stored in your private Library.

Sample output

Official Wan 2.5 demo · video

Reference rendering shipped by Alibaba — same endpoint you'll run inside Lumineer.

Wan 2.5
video

Specs

Technical characteristics of Wan 2.5 as it runs inside Lumineer.

VendorAlibaba
Categoryimage → video
Tierpremium
Price per run188 credits (~$1.50)
Typical latency~120s
Aspect ratios16:9 · 9:16 · 1:1
Resolutions1080p · 720p · 480p
Durations5s · 10s
Referencessingle
Seed controlNo
Negative promptNo
API accessREST via Lumineer (no fal.ai key required)

Related models