Wan 3.0
alibaba/wan-3.0Commercial useWan 3.0 is Alibaba's text to video model, available here through Lithovas's unified API. It generates a short video clip from a written description. It takes a text prompt as input. Requests run server-side with your Lithovas credit balance, so you get Alibaba quality without managing a separate Alibaba account or key.
Text prompt describing the video to generate. Required for normal text-to-video requests; up to 20,000 characters.
Optional public image URL used as the exact first frame. OpenRouter Wan 3.0 currently supports first-frame control, not last-frame control.
Reference image URLs used as visual guidance via OpenRouter input_references.
Output duration in seconds. Wan 3.0 supports 2–30 seconds.
Output video resolution tier.
Output aspect ratio.
Whether to generate audio alongside the video.
Random seed from 0 to 2147483647. Determinism depends on upstream support.
