Vidu Q2 reference-to-video AI generator

Vidu Q2 is a video model from ShengShu Technology, the company behind Vidu. Its reference-to-video mode, launched in October 2025, builds each clip from 1 to 7 reference images and keeps the characters, props and places in them consistent.

Vidu Q2 builds each clip directly from 1 to 7 reference images and keeps the people, props and places in them consistent as the clip plays. On DreamBench those images are the character, prop and location sheets it designs for your video, built from any pictures you upload. It doesn’t start from a single opening frame, so it isn’t image-to-video in the usual sense: the sheets carry each subject’s look from scene to scene.

No. ShengShu has since released Vidu Q3 and lists a Vidu Q4 Preview in its API. DreamBench runs Vidu Q2 in its reference-to-video mode because it builds every clip straight from the sheets of your cast, props and places, which keeps them consistent from scene to scene.

No. DreamBench runs Vidu Q2 itself. Vidu offers Q2 Pro on its own API, where it can also take reference videos and edit footage.

Each clip Vidu Q2 generates can run up to 10 seconds. A whole DreamBench video can be longer, up to 5 minutes depending on your plan, because DreamBench plans the scenes and joins the clips into one video.

Vidu Q2 renders at 720p or 1080p on DreamBench. You can make it square (1:1), widescreen (16:9), portrait (9:16) or 4:3.

Only background music. On DreamBench, Vidu Q2’s sound switch asks Vidu to add music it chooses to suit the clip, with no dialogue, voices or sound effects, and Vidu leaves music out of clips that run 9 or 10 seconds, so those play silent even with the switch on. Tell the story through faces and gestures.

Vidu Q2 uses credits for each second of video, at a rate that depends on the resolution and stays the same with the music on or off, plus a base amount per video. The table on this page shows the current rates, and you always see the exact cost before you generate.

Yes. On DreamBench, each Vidu Q2 video uses credits, which come with a paid plan, and Vidu Q2 is on every plan. A second costs the same with the music on or off, and the composer shows the exact cost before you generate.

Name each character, prop and place as it appears on its sheet, describe the action through faces and gestures, and don’t write dialogue: on DreamBench, Vidu Q2 adds background music at most, never speech.

Both are on DreamBench, so run the same brief on each and keep the one you prefer. Vidu Q2 builds each clip of up to 10 seconds straight from the sheets of your cast, props and places, with background music at most; the Seedance 2.0 page, linked below, states its own limits.

No. DreamBench offers Vidu Q2 through its composer, in the browser on a phone, tablet or computer, with nothing to install. Developers can use Vidu’s own API, linked under Sources, where the prices and limits are ShengShu’s, not DreamBench’s.

Videos made on the Starter plan carry a small DreamBench watermark; Plus, Ultra and Studio remove it. The picture is otherwise as Vidu Q2 generated it.

Yes. DreamBench’s Terms of Use allow personal and commercial use, including ads and client work. You need the rights to anything you upload.

Available on every plan. Video length depends on the plan: up to 30 seconds on Starter, 2 minutes on Plus and 5 minutes on Ultra and Studio.

4: 1:1, 16:9, 9:16 and 4:3. Pick the ratio in the composer, or name it in the prompt and DreamBench sets it for you.

Press Create with Vidu Q2, or pick it in the composer, and describe the video you have in mind. DreamBench writes the scene plan, generates every scene with Vidu Q2 and joins them into one video. Then give any scene new direction and keep the rest.