☁️BizyAir Vidu Q1 Text To Video

Text-to-video from Vidu Q1, no source image needed. Same family as the other two Vidu nodes in this pack - one animates from a first frame, one animates from a set of reference images, this one generates purely from a prompt. Vidu, like Kling and Hailuo, never released local weights, so this is the only route to it from ComfyUI: your BizyAir account calls out to Vidu's API and a clip comes back.

The thing that actually sets this node apart from its two siblings isn't the lack of an image input - it's a dedicated style toggle that neither of the other Vidu nodes exposes.

What it does

Prompt goes in, video comes out, entirely generated on Vidu's servers. No local GPU work, no downloaded checkpoint - you're spending BizyAir credits per generation, same as every other API node in this pack.

Inputs and outputs that matter

• prompt - multiline text describing the scene. With no image to anchor composition, this is carrying the full load; be specific about subject, setting, and camera behavior rather than relying on adjectives.
• seed - set a value for reproducibility, leave it for variation.
• style - general or anime. This is unique to the text-to-video node; neither of the image-conditioned Vidu nodes in this pack exposes it. If you're going for an anime look, flip this switch instead of fighting your prompt wording to get there - it's a real model-level toggle, not a prompt trick.
• duration - currently a single option, 5 seconds.
• aspect_ratio - 16:9, 9:16, or 1:1. Three choices here, more than either of the other two Vidu nodes offer (they're locked to whatever the reference/first-frame image implies).
• resolution - currently fixed to 1080p.
• movement_amplitude - auto / small / medium / large, controlling how much motion the model puts into the scene. Turn it down for something closer to a slow pan, up for more energetic motion.
• bgm - auto-generates and adds a matching background track, no length limit, adapted to the clip automatically. Same feature every Vidu node in this pack carries and no Veo or Wan node does.

Output is video and bizyair_model_name, an internal label rather than something to key logic off of.

How to install it

• ComfyUI Manager - search "BizyAir", install, restart.
• Manual - cd ComfyUI/custom_nodes && git clone https://github.com/siliconflow/BizyAir.git, restart.
• Comfy CLI - comfy node install bizyair.

Register and set your API key before your first run - the in-app prompt or bizyair.siliconflow.cn directly. Nothing in this pack works without one, and every generation here is metered.

Common issues & troubleshooting

The style: anime toggle didn't change much. It's a coarse setting, not a full art-direction system - you'll still get better results describing the specific anime aesthetic you want (flat shading, cel-shaded, particular era) in the prompt itself rather than relying on the switch alone.

A prompt with no image gives you a composition you didn't expect. Normal for pure text-to-video - there's more interpretive freedom without a starting frame to anchor the shot. Explicit camera language (wide shot, close-up, static vs. panning) narrows the result more reliably than piling on descriptive adjectives.

BizyAir appeared in your setup without you installing it deliberately. A known ComfyUI Manager bug used to auto-resolve missing nodes onto BizyAir incorrectly, which is behind the handful of "is this malware" threads floating around. It's a real commercial pack, not malicious - remove it through the Manager if you don't want the dependency, keep it if you're actually using nodes like this one.

Queue does nothing. Check the API key and credit balance first; there's no offline path for this node.