zhipu modelsimage generation api

zhipu/glm-image/edit

Name: Glm Image
Author: zhipu

Glm Image Edit by zhipu - AI-powered image editing, style transfer, and transformation. Edit photos with natural language instructions, remove backgrounds, change styles, and enhance images effortlessly.

Input

Prompt(string)

Text prompt for image generation.

Image URL or upload(array)

Upload

PNG, JPEG, WebP, or GIF · 20 MiB maximum each

URL(s) of the condition image(s) for image-to-image generation. Supports up to 4 URLs for multi-image references.

Guidance Scale(number)

110

Classifier-free guidance scale. Higher values make the model follow the prompt more closely. Range: 1 to 10.

Num Inference Steps(integer)

10100

Number of diffusion denoising steps. More steps generally produce higher quality images. Range: 10 to 100.

Output Format(string)

Output image format. Allowed values: jpeg, png.

Aspect Ratio(string)

The aspect ratio of the generated image. Allowed values: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16.

Seed(integer)

Random seed for reproducibility. The same seed with the same prompt will produce the same image.

Idle

Example output — click Run to generate your own

Production endpoint

Send your first request

OpenAI-compatible endpoint with unified authentication and usage tracking.

POSThttps://api.sandbase.ai/v1/run

Model IDzhipu/glm-image/edit

Input Schema

7 parameters · 1 required · 6 optional

Parameter	Type	Required	Description
`prompt`	string	Required	Text prompt for image generation.
`seed`	integer	Optional	Random seed for reproducibility. The same seed with the same prompt will produce the same image.
`images`	string[]	Optional	URL(s) of the condition image(s) for image-to-image generation. Supports up to 4 URLs for multi-image references.
`aspect_ratio`	string	Optional	The aspect ratio of the generated image. · Options: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 21:916:93:24:35:41:14:53:42:39:16
`output_format`	string	Optional	Output image format. · Options: jpeg, png · Default: "jpeg" jpegpng
`guidance_scale`	number	Optional	Classifier-free guidance scale. Higher values make the model follow the prompt more closely. · Min: 1 · Max: 10 · Default: 1.5
`num_inference_steps`	integer	Optional	Number of diffusion denoising steps. More steps generally produce higher quality images. · Min: 10 · Max: 100 · Default: 30

Output Schema

Field	Type	Description
`id`	string	Unique identifier for the generation task
`status`	string	Task status: pending, running, completed, failed, timeout
`model`	string	Model used for the generation
`outputs`	array	Array of output items
`outputs[].url`	string	URL of the generated artifact
`outputs[].content_type`	string	MIME type (e.g. image/png, video/mp4)
`error`	object \| null	Error details if failed, null on success
`error.type`	string	Machine-readable error type code
`error.message`	string	Human-readable error description

Async Workflow

This model uses asynchronous execution. Submit a request and poll for the result.

Submit — POST to /v1/run, receive an id
Poll — GET /v1/run/{id} until status is completed, failed, or timeout
Retrieve — Read outputs from the completed response

Code Examples

Ready-to-run snippets

# Step 1: Submit
curl -X POST https://api.sandbase.ai/v1/run \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
  "model": "zhipu/glm-image/edit",
  "images": [
    "https://static.sandbase.ai/examples/zhipu/glm-image/edit/input_images_0.png"
  ],
  "prompt": "Make the dress red.",
  "output_format": "jpeg",
  "guidance_scale": 1.5,
  "num_inference_steps": 30
}'

# Step 2: Poll result (replace <id>)
curl https://api.sandbase.ai/v1/run/<id> \
  -H "Authorization: Bearer YOUR_API_KEY"

← Back to Glm Image

API README

GLM Image Edit

GLM Image Edit is a specialized SandBase endpoint built to revise one or more condition images through text while retaining direct control over generation strength and refinement depth. It is most useful when teams need multi-image conditioning, deterministic iteration, adjustable prompt guidance, and selectable denoising effort, with the route’s inputs defining a repeatable production contract instead of leaving critical delivery choices to an ad-hoc manual workflow.

Use this exact route when its input mode matches the creative asset already in hand: describe the visual or spoken result clearly, set only the controls that support the intended delivery, and keep the subject, action, environment, camera or performance direction internally consistent. The result is returned asynchronously as a downloadable media URL suitable for review, automation, or downstream finishing.

Highlights

Multi-image conditioning. Supply as many as four visual references when an edit must reconcile subject, style, product, or layout cues from several sources.
Reproducible iterations. Reuse a seed with the same inputs to revisit a promising result and compare controlled prompt adjustments.
Adjustable prompt adherence. Guidance scale from 1 to 10 lets teams decide how strongly the written instruction should steer the revised image.
Refinement-depth control. Choose 10–100 inference steps to trade faster exploration for a more thoroughly resolved generation pass.

Pricing

Each image-edit request costs $0.0500.

Operation	Price
GLM Image edit	$0.0500

When to Use

Scenario	Why this route fits
Concept development	Choose GLM Image Edit when its edit workflow matches the starting material and you need several clearly directed variations.
Production iteration	Use explicit duration, resolution, ratio, seed, or quality controls to compare versions without changing the core creative brief.
Channel adaptation	Generate directly in the landscape, square, portrait, or vertical format required by the destination whenever that control is available.
Automated pipelines	Integrate the asynchronous media URL into review queues, asset libraries, publishing tools, or a later finishing stage.
Alternative route	Pick a related text-, image-, reference-, edit-, or turbo route when the available source media or required degree of control is different.

Prompt Guide

Lead with the main subject or source asset, then describe the intended action or transformation, environment, composition, camera or vocal delivery, lighting and mood. Keep instructions concrete and compatible; use the route’s explicit fields for duration, resolution, ratio, quality, voice, or reproducibility instead of burying those settings in prose.

{
  "prompt": "A cinematic product reveal with deliberate subject motion, coherent lighting, and a slow camera push.",
  "aspect_ratio": "21:9",
  "output_format": "jpeg",
  "guidance_scale": 1.5,
  "num_inference_steps": 30,
  "images": [
    "https://example.com/reference-1.jpg"
  ]
}

Technical Specs

Property	Details
Model ID	`zhipu/glm-image/edit`
Required inputs	`prompt`
Execution	Asynchronous; poll the returned generation ID
Output	Downloadable media URL
`seed`	integer; optional
`images`	array; optional
`prompt`	string; required
`aspect_ratio`	string; optional; choices: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16
`output_format`	string; optional; choices: jpeg, png; default: jpeg
`guidance_scale`	number; optional; range: 1–10; default: 1.5
`num_inference_steps`	integer; optional; range: 10–100; default: 30

Related Models

Model	Best for
`zhipu/glm-image`	Alternative glm image workflow
`bfl/flux-2/dev`	Alternative image generation or editing workflow
`google/nano-banana-pro`	Alternative image generation or editing workflow
`bytedance/seedream/4.5/edit`	Alternative image generation or editing workflow