Aliyun Bailian | 阿里云百炼Install

Summary

Generate or edit images using Aliyun Bailian Wan 2.7 AI models with customizable instructions, input images, output count, resolution, and advanced options for consistent image sets, watermarking, and deterministic seeds.

Inputs

• Model | 模型 (required) — Choose between Wan 2.7 Image Pro (supports 4K text-to-image) and Wan 2.7 Image standard version (faster generation).
• Instruction | 指令 (required) — Text prompt describing the image generation or editing request, including subject, style, and constraints.
• Input Images | 输入图片 URL 或 Base64 列表 — Optional list of 0 to 9 images (URLs or Base64) to use as input references or for image editing.
• Sequential Output | 组图输出 — Boolean to enable generation of a consistent image set sharing subject and style rather than independent single images.
• Image Count | 生成图片上限 — Maximum number of images to generate (1-4 if sequential output off, 1-12 if enabled). Actual returned count may be fewer.
• Size Name or ID | 分辨率档位或尺寸 — Select resolution tier (1K, 2K, 4K) or specify custom dimensions. 4K only supported by Pro model for pure text-to-image without sequential output.
• Negative Prompt | 反向提示词 — Optional negative prompt to suppress undesired content in generated images.
• BBox List JSON | 交互式编辑框选 JSON — Optional JSON array defining interactive editing bounding boxes aligned with input images for targeted edits.
• Color Palette JSON | 颜色主题 JSON — Optional JSON defining color palettes to guide image colors; effective only when sequential output is disabled.
• Advanced Options | 高级选项 — Additional settings including Thinking Mode (improves quality but increases latency), watermark inclusion, fixed or randomized seed for deterministic outputs.

Output shape

A list of generated or edited image data objects, with count up to user-specified limits and consistent styling if sequential output enabled.

Image count limits vary depending on sequential output setting. 4K resolution generation is limited to the Pro model without input images or sequential output. The node may poll for task completion within configurable intervals and maximum wait times.

Examples

Example 1: Generate a single cinematic style product poster at 2K resolution using the Pro model and default settings, with no input image.

Set model to wan2.7-image-pro, instruction describing the cinematic poster, enable no sequential output, image count 1, size to 2K, and default advanced options.

Example 2: Edit up to 3 input images with consistent style and subject, at 1K resolution, adding a color palette constraint and watermark.

Provide up to 3 images as input, set Sequential Output to true, Image Count to 3, Size to 1K, specify a Color Palette JSON, enable watermark in Advanced Options.

Links

• n8n community nodes installation guide