Nodes/BizyAirPlus/X AI Grok Imagine Video Text To Video (Official Edition)
ComfyUI Node

X AI Grok Imagine Video Text To Video (Official Edition)

Grok Imagine text-to-video through xAI's own API

By siliconflow·Created 5 months ago·Updated about a month ago· 20
X AI Grok Imagine Video Text To Video (Official Edition)
  • api_config
  • video
  • request_id
  • urls
prompt
resolution720p
aspect_ratio16:9
duration6
skip_errorfalse

The official edition of Grok Imagine text-to-video. The distinction from the channel edition is plumbing rather than model: this node targets xAI's official API surface instead of SiliconFlow's BizyAir channel. For a beginner the practical question is which key you have - if you're already set up with xAI, this is your node; if you want one SiliconFlow key to cover a pile of models, the channel version is the easier door in.

Functionally it's a sentence in, video out. Prompt, resolution, aspect ratio, duration - done. The photorealistic tier Grok Imagine competes in (Veo, Kling, the current video frontier) is exactly where local video setups stop being a hobby and start being a profession, so for most people the hosted route is not a compromise, it's the sane option. And the official edition keeps that simple because the model does the heavy lifting.

How it works

Same BizyTRD wrapper pattern as every node in this pack - BizyAirPlus ON, request out, VIDEO tensor back. The controls: prompt (required), resolution (required, default 720p), aspect_ratio (required, default 16:9), and duration as a combo with a default of 6 seconds. Unlike the channel edition's 6–30 second slider, official durations come from a fixed list - so if you want a specific unusual length, the channel node is the one with the slider.

api_config lets you override the endpoint and key for a single request, which on an official edition is the natural place to put your vendor key if it differs from your BizyAir default. skip_error, when enabled, turns a failed request into a placeholder output instead of aborting the whole workflow.

Inputs and outputs

  • prompt (required, multiline) - the whole creative input.
  • resolution (default 720p), aspect_ratio (default 16:9), duration (combo, default 6).
  • Outputs: video (VIDEO), request_id, urls.

Installing BizyAirPlus

Search "BizyAirPlus" in ComfyUI Manager, or:

cd /path/to/ComfyUI/custom_nodes
git clone https://github.com/siliconflow/BizyAirPlus.git
cd BizyAirPlus
python -m pip install -r requirements.txt

Restart, get an API key at bizyair.ai, flip the BizyAirPlus button ON. No model downloads; just bizyair-cloudberry and bizytrd as pip dependencies.

Common issues

The standard pack trio - button missing (wrong dependency environment, pip show bizyair-cloudberry), cloud failures (key and network, then the console), and metering (video is the priciest thing here). The official-edition wrinkle: if auth fails while the channel nodes work fine, you're almost certainly pointing the wrong key or base_url at the official endpoint - that's exactly what api_config is for.

CategoryBizyAirModelServices/BizyTRD/Grok Video

Inputs (6)

NameTypeDefaultDescription
promptSTRINGPrompt words for video generation
resolutionCOMBO720pVideo resolution
aspect_ratioCOMBO16:9Aspect ratio of video
durationCOMBO6Video length in seconds
api_configoptBIZYAIR_OPENAPI_CONFIG单次请求覆盖base_url和api_key
skip_erroroptBOOLEANfalse开启后遇到错误不中断工作流,输出对应类型的错误占位符

Outputs (3)

NameTypeDescription
videoVIDEO
request_idSTRING
urlsSTRING