Aliyun Bailian | 阿里云百炼Install

Summary

Generate a video clip from a detailed natural language prompt using the Aliyun Bailian text-to-video models.

Inputs

• Model (required) — Select the Aliyun Bailian text-to-video model version to use for generation.
• Prompt | 文本提示词 (required) — Detailed natural-language prompt describing the desired video content.
• Negative Prompt | 反向提示词 — Optional instructions listing elements to avoid in the generated clip.
• Resolution | 分辨率档位 — Resolution tier selection, applicable only for the wan2.7-t2v model.
• Ratio | 宽高比 — Output aspect ratio, applicable only for the wan2.7-t2v model.
• Size Name or ID — Output video resolution in width*height format; varies by model version.
• Duration (Seconds) (required) — Requested length of the generated video clip, with allowed values depending on the model version.
• Auto Enhance Prompt — Toggle to enable Aliyun Bailian's prompt optimization for richer scene details.
• Shot Type | 镜头模式 — Select between single or multi-shot mode; effective only on wan2.6 model when prompt optimization is enabled.
• Generate Audio — Enable automatic audio dubbing when supported by the selected model.
• Audio URL — Optional public URL to a background audio track, supported by select models.
• Advanced Options | 高级选项 — Additional parameters including embedding an AI watermark, specifying a deterministic seed for reproducibility, and whether to randomize the seed at runtime.
• Wait for Task Completion — Whether to poll Aliyun Bailian until the video generation task completes or fails, or return immediately with the task ID.
• Poll Interval (MS) — Interval between polling attempts, used if waiting for task completion.
• Max Wait (Seconds) — Maximum wait time before polling times out, used if waiting for task completion.

Output shape

an object representing the video generation task status and final result URL(s) or media metadata

If 'Wait for Task Completion' is enabled, the node polls until the generation completes or fails, returning final output data. Otherwise, it returns the task ID immediately for asynchronous checking. Output fields vary per model and options selected, including video URLs, status, and optional audio track data.

Examples

Example 1: Create a 10-second 1080p video with a prompt describing a cute cat on a sofa, using the wan2.7-t2v model at 16:9 aspect ratio, with auto prompt enhancement enabled, waiting for completion to get the video URL.

Set Model to wan2.7-t2v; set Prompt with detailed description; select Resolution 1080P and Ratio 16:9; set Duration to 10; enable Auto Enhance Prompt; enable Wait for Task Completion; optionally specify Poll Interval and Max Wait.

Links

• n8n community nodes installation guide