vidoro

AI model transparency

AI models used by Vidoro

Every AI model behind Vidoro's generators, in one place. Each generator's Model information section reads from this same registry, so the site never claims a model it does not call.

Last updated: 2026-10-11

Template

Hotel Lobby AI

Two portraits, one fixed 15-second orange-studio performance.

Open Hotel Lobby AI
  1. STEP 1 — Starting frame

    Wan 2.6 Image

    Model ID: wan2.6-imageProvider: Alibaba Cloud Model Studio (DashScope)

    A multimodal image-edit call. It replaces the identities of the template's two existing performers and keeps the orange studio, poses, microphone, camera and lighting unchanged, so the motion stage starts from a predictable composition.

    Output: One 1280×720 (16:9) frame.

  2. STEP 2 — Video generation

    Wan 2.6 Image-to-Video

    Model ID: wan2.6-i2vProvider: Alibaba Cloud Model Studio (DashScope)

    An image-to-video call driven by the template's fixed motion prompt. The choreography, camera behaviour and timing come from the template, not from a user prompt.

    Output: 15-second 16:9 video at 720P, silent, no watermark.

Template

Raindance AI

Two portraits, one directed 15-second vertical performance.

Open Raindance AI
  1. STEP 1 — Video generation

    Wan 3.0 Video

    Model ID: wan3.0-videoProvider: Alibaba Cloud Model Studio (DashScope)

    A reference-guided video call. Your two photos supply the character identities and the fixed reference video supplies the scene changes, framing and motion, so every run follows the same directed sequence.

    Output: 15-second 9:16 vertical video at 720P or 1080P, with audio, no watermark.

FAQ

Model questions, answered.

Does Vidoro build its own AI models?

No. Generation runs on hosted models accessed through Alibaba Cloud Model Studio (DashScope). Vidoro builds and maintains the fixed templates, prompts, reference material, identity handling and task orchestration on top of those models.

Why does this page exist?

Every generator on Vidoro shows a short Model information section, and all of those sections read from the single registry on this page. When the pipeline changes, the registry is updated once and every generator page stays accurate automatically.

Can the models change?

Yes. Vidoro may switch a stage to a newer model version when a provider release improves the fixed templates. The list on this page, including its last-updated date, always reflects the models actually being called.

Are the generated videos watermarked?

No. Vidoro requests generation without a provider watermark on both current templates, so paid exports carry no Vidoro or provider watermark.