Generate Image — AI image generation, 20+ models

Tools / Generate Image

Text to Image

Generate an image from a text prompt. 20+ models with different quality, speed, style, and cost tradeoffs. Returns image URL(s) and downloadable assets.

Returns: Image URL(s), downloadable asset(s) via asset system, model used, seed, dimensions, and request metadata

Parameters

prompt *stringDetailed text description of the image to generate. Be specific about subject, style, lighting, composition, and mood.

modelstringModel to use — registry key, name, or endpoint ID. Use list_models for options. Omit for default.

image_sizestringSize preset: square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3 (default), landscape_16_9.

aspect_ratiostringAspect ratio (used by some models instead of image_size). Common aliases like "square", "landscape", "portrait" are also accepted.

widthnumberCustom width in pixels (use with height instead of image_size).

heightnumberCustom height in pixels (use with width instead of image_size).

num_imagesnumberNumber of images to generate (1-4). Defaults to 1.

seednumberRandom seed for reproducible results. Same seed + prompt → similar output.

output_formatstringOutput image format. Defaults to jpeg.

negative_promptstringThings to avoid in the image (e.g. "blurry, low quality, text"). Not all models support this.

guidance_scalenumberHow closely to follow the prompt (1-20). Higher = more literal. Not all models support this.

num_inference_stepsnumberNumber of denoising steps. More steps = higher quality but slower. Not all models support this.

stylestringStyle preset (model-specific). Values depend on model — use list_models to see supported styles.

style_presetstringDetailed style preset (model-specific, e.g. "WATERCOLOR", "POP_ART", "80S_ILLUSTRATION"). Use list_models to see which models support this.

rendering_speedstringRendering speed — affects quality and cost. Only supported by select models.

qualitystringQuality tier — affects cost. Only supported by select models.

backgroundstringBackground mode. Only supported by select models.

enhance_promptbooleanLet the model expand/improve your prompt. Only supported by select models.

rawbooleanLess processed, more natural-looking output. Only supported by select models.

colorsarrayColor palette as [{r,g,b}] objects to guide generation. Only supported by select models.

lorasarrayLoRA weights as [{path, scale}] objects. Only supported by select models.

safety_tolerancestringSafety level (1-6). Higher = more permissive. Only supported by select models.

extra_paramsobjectAdditional model-specific parameters passed directly to the inference API. Use for custom models or advanced features not listed above.

Example

Generate a product photo with the default model

curl -H "Authorization: Bearer $TOOLROUTER_API_KEY" \
  -d '{
  "tool": "generate-image",
  "skill": "text_to_image",
  "input": {
    "prompt": "A sleek wireless headphone on a marble surface, soft studio lighting, product photography, 4K"
  }
}' \
  https://api.toolrouter.com/v1/tools/call

Edit Image

Edit, transform, or inpaint an existing image using AI. Provide an image URL and a text prompt describing the edit. Supports style transfer, object removal/addition, background changes, and targeted edits with masks.

Returns: Edited image URL(s), downloadable asset(s), source image URL, model used, seed, and request metadata

Parameters

image_url *stringURL of the source image to edit. Must be publicly accessible.

promptstringText description of the edit to make (e.g. "change the background to a beach sunset", "remove the person on the left").

modelstringModel to use for editing. Use list_models with capability "editing" or "inpainting" to discover options. Omit to use the default edit model.

mask_urlstringURL of mask image for inpainting (white = area to edit). Required for models with "inpainting" capability.

strengthnumberHow much to change the image (0-1). Lower = closer to original. Supported by image-to-image models.

image_sizestringSize preset: square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3 (default), landscape_16_9.

aspect_ratiostringAspect ratio (used by some models instead of image_size). Common aliases like "square", "landscape", "portrait" are also accepted.

widthnumberCustom width in pixels.

heightnumberCustom height in pixels.

num_imagesnumberNumber of images to generate (1-4). Defaults to 1.

seednumberRandom seed for reproducible results. Same seed + prompt → similar output.

output_formatstringOutput image format. Defaults to jpeg.

negative_promptstringThings to avoid in the image (e.g. "blurry, low quality, text"). Not all models support this.

guidance_scalenumberHow closely to follow the prompt (1-20).

num_inference_stepsnumberNumber of denoising steps.

enhance_promptbooleanLet the model expand your prompt.

extra_paramsobjectAdditional model-specific parameters passed directly to the inference API.

Example

Edit an image with natural language using Kontext

curl -H "Authorization: Bearer $TOOLROUTER_API_KEY" \
  -d '{
  "tool": "generate-image",
  "skill": "edit_image",
  "input": {
    "image_url": "https://placehold.co/800x600.png",
    "prompt": "Add a decorative golden frame border around the entire image"
  }
}' \
  https://api.toolrouter.com/v1/tools/call

Upscale Image

AI-upscale an image up to 10x resolution. Enhances detail and sharpness while maintaining quality. Choose a target resolution (720p-4K) or a scale factor (1-10x).

Returns: Upscaled image URL, downloadable asset, upscale factor or target resolution, and request metadata

Parameters

image_url *stringURL of the image to upscale. Must be publicly accessible.

upscale_factornumberScale factor (1-10). Default 2x. Ignored if target_resolution is set.

target_resolutionstringTarget resolution instead of scale factor.

seednumberRandom seed for reproducible results. Same seed + prompt → similar output.

output_formatstringOutput image format. Defaults to jpeg.

Example

Upscale an image 2x

curl -H "Authorization: Bearer $TOOLROUTER_API_KEY" \
  -d '{
  "tool": "generate-image",
  "skill": "upscale_image",
  "input": {
    "image_url": "https://placehold.co/400x300.png"
  }
}' \
  https://api.toolrouter.com/v1/tools/call

List Models

List all available image generation models with capabilities, pricing, supported parameters, and descriptions. Filter by capability type. Models not in the registry can still be used by passing their model endpoint ID directly.

Returns: Array of model objects with key, capabilities, pricing, supported parameters, and descriptions

Parameters

capabilitystringFilter models by capability. Omit to show all.

Example

List all available image models

curl -H "Authorization: Bearer $TOOLROUTER_API_KEY" \
  -d '{
  "tool": "generate-image",
  "skill": "list_models",
  "input": {}
}' \
  https://api.toolrouter.com/v1/tools/call

Quick Start

MCP (Claude Code)

claude mcp add --transport stdio \
  --env TOOLROUTER_API_KEY=YOUR_API_KEY \
  toolrouter -- npx -y toolrouter-mcp

REST API

curl -H "Authorization: Bearer $TOOLROUTER_API_KEY" \
  -d '{"tool":"generate-image","skill":"text_to_image","input":{}}' \
  https://api.toolrouter.com/v1/tools/call

Use Cases

Open Create Marketing Banners

Create Marketing Banners

Generate professional marketing banners for campaigns, promotions, and seasonal sales with AI image generation.

Generate Image4 agent guides

Open Generate Product Mockups

Generate Product Mockups

Create realistic product mockups and lifestyle shots without expensive photography or 3D rendering.

Generate Image4 agent guides

Open Dub Marketing Videos

Dub Marketing Videos

Translate and dub your marketing videos into multiple languages to reach international audiences.

Audio Dubber4 agent guides

Open Remove Background Noise from Recordings

Remove Background Noise from Recordings

Clean up recordings by removing background noise, hum, wind, and environmental sounds.

Audio Isolator4 agent guides

View all use cases for Generate Image→

Workflows

Open Content Marketing Pipeline

Content Marketing Pipeline

Build a complete content creation pipeline from keyword research through deep topic research to visual asset generation.

Keyword Research

Web Search

Deep Research

Generate Image

4 steps4 tools

Open Full-Funnel Ad Research

Full-Funnel Ad Research

Research competitor advertising across platforms, analyze their strategy, study their organic content, and generate your own ad creatives.

Ad Library Search

Competitor Research

Social Media Content

Generate Image

4 steps4 tools

Open Landing Page Optimizer

Landing Page Optimizer

Optimize landing pages by auditing SEO elements, testing performance, capturing responsive screenshots, and generating new hero visuals.

SEO Analysis

Lighthouse Audit

Web Screenshot

Generate Image

4 steps4 tools

Open Social Media Growth Kit

Social Media Growth Kit

Accelerate social media growth by auditing your profiles, discovering trends, analyzing top content, and generating visuals.

Social Profiles

Frequently Asked Questions

Can I create an image from a text prompt?

Yes. `text_to_image` turns a prompt into a new image and returns downloadable image assets.

Can I edit an image I already have?

Yes. `edit_image` can change backgrounds, add or remove objects, and make targeted visual edits.

Can I upscale an image for higher resolution?

Yes. `upscale_image` can enhance an image up to 10x resolution, including 4K-level output.

How do I know which model to use?

Use `list_models` to compare capabilities, speed, and style strengths before you generate.