Wan AI Video Generator

Wan AI can turn a single image into a short video when the selected entry accepts that input. Reviral's currently priced Wan entries are Wan 2.6 and Wan 2.6 Flash. Standard accepts one reference image, both expose a reference-video path, and the family offers 5-, 10-, or 15-second clips at 720p or 1080p with an audio-generation switch.

Some searches call this a Wan 2.5 video generator, but that is not the name of the current priced registry entry. Choose Wan 2.6 or Flash in the selector and describe the actual output with that label. The wider registry also contains Wan 3.0 routes with different reference and duration controls, but those routes are not used in this page's customer price table.

Wan 2.6 and Wan 2.6 Flash offer three fixed lengths: five, ten, or fifteen seconds. Five seconds is useful for a motion proof or insert, ten for one fuller beat, and fifteen for a compact scene with setup and resolution. A fixed list prevents arbitrary timing, so design the action around one of those durations instead of writing a timeline the model form cannot accept.

Both priced entries list 720p and 1080p. Standard explicitly lists 16:9, 9:16, 1:1, 4:3, and 3:4, covering widescreen, vertical, square, and common portrait formats. Flash uses the entry's current default aspect behavior, so verify the form rather than copying Standard's aspect list. Resolution and aspect are delivery constraints, not quality rankings.

Standard lists one reference image and a reference-video field. Flash marks reference video as an alternative input. Both include a generate-audio switch whose default is true in the registry. Their verified cost rows also include a lower cost multiplier when audio is disabled, though the exact customer price remains a live quote. Decide whether sound belongs in the source clip and review synchronization after generation.

Identify what in the image must stay recognizable, then describe one action and the camera in order. For a product, specify whether it remains fixed while the camera moves or whether a hand lifts, turns, or places it. For a character, state body direction, expression change, and ending pose. The prompt supplies motion that a still image cannot.

Choose the duration from the action. Five seconds can cover a quick reveal or small gesture. Ten seconds allows a more readable move and landing. Fifteen seconds can carry a brief setup and finish, but too many unrelated changes still create an ambiguous target. Split multiple locations, claims, or camera setups into separate shots and plan the final sequence deliberately.

If audio stays enabled, write the short sound requirement separately: a line, ambience, or effect. Do not overload a fifteen-second prompt with a long script. If audio is disabled, plan voiceover, captions, music, and effects in the edit. Record whether the generation used Standard or Flash so later reviewers can reproduce the same control set and pricing context.