Image Generation

Generate images from text prompts across 5 providers with a single unified API.

Overview
generateImage() API
Provider Reference
Provider Options Passthrough
Local Setup (A1111 and ComfyUI)
Custom Image Provider
Usage Tracking

Overview

generateImage() is a provider-agnostic factory. Call it the same way regardless of which backend you use — switch providers by changing provider without touching application code.

Supported providers:

Provider	ID	Key Env Var	Best For
OpenAI	`openai`	`OPENAI_API_KEY`	Photorealistic, DALL-E quality
Stability AI	`stability`	`STABILITY_API_KEY`	Art styles, SDXL models
Replicate	`replicate`	`REPLICATE_API_TOKEN`	FLUX, community models
OpenRouter	`openrouter`	`OPENROUTER_API_KEY`	Model routing, fallback
Stable Diffusion Local	`stable-diffusion-local`	`STABLE_DIFFUSION_LOCAL_BASE_URL`	Privacy, no API cost
Ollama	`ollama`	`OLLAMA_BASE_URL`	Local, offline

Mature content: pass policyTier: 'mature' or 'private-adult' and generateImage reroutes through the uncensored catalog (Replicate community models, face-consistency aware) and disables the provider safety filter automatically. See UNCENSORED_CONTENT.md for the full API and catalog.

generateImage() API

import { generateImage } from '@framers/agentos';

const result = await generateImage({
  // Required
  provider: 'openai',
  prompt:   'A futuristic city at sunset with flying cars and neon lights, photorealistic.',

  // Optional
  model:          'gpt-image-1',      // provider default if omitted
  negativePrompt: 'blurry, text, watermark',
  width:          1024,
  height:         1024,
  aspectRatio:    '16:9',             // alternative to width/height
  numImages:      1,
  outputFormat:   'png',              // 'png' | 'jpeg' | 'webp'
  quality:        'hd',               // provider-specific quality tier
  seed:           42,                 // for reproducibility
  steps:          30,                 // diffusion steps (local/Stability)

  // Provider-specific options (see per-provider sections)
  providerOptions: {},
});

// Result shape
console.log(result.images[0].url);       // URL or base64 data URI
console.log(result.images[0].mimeType);  // 'image/png'
console.log(result.images[0].width);     // actual width
console.log(result.images[0].height);    // actual height
console.log(result.modelId);             // model used
console.log(result.providerId);          // provider used
console.log(result.usage.totalImages);   // 1

Common Parameters

Parameter	Type	Description
`provider`	`string`	Provider ID (required)
`prompt`	`string`	Text description of the image (required)
`model`	`string`	Model ID — uses provider default if omitted
`negativePrompt`	`string`	What to exclude (Stability, Replicate, local)
`width`	`number`	Image width in pixels
`height`	`number`	Image height in pixels
`aspectRatio`	`string`	`'1:1'`, `'16:9'`, `'9:16'`, `'4:3'`, etc.
`numImages`	`number`	Number of images to generate (default: 1)
`outputFormat`	`string`	`'png'` (default), `'jpeg'`, `'webp'`
`quality`	`string`	Quality tier (`'standard'`, `'hd'`)
`seed`	`number`	Seed for reproducible generation
`steps`	`number`	Diffusion steps (local models, default: 20–50)

Provider Reference

OpenAI (gpt-image-1)

const image = await generateImage({
  provider: 'openai',
  model:    'gpt-image-1',  // or 'dall-e-3', 'dall-e-2'
  prompt:   'A minimal logo for a tech startup, flat design, blue and white.',
  quality:  'hd',
  size:     '1024x1024',    // OpenAI uses size instead of width/height
  providerOptions: {
    openai: {
      style:            'vivid',    // 'vivid' | 'natural'
      responseFormat:   'url',      // 'url' | 'b64_json'
    },
  },
});

Default model: gpt-image-1 Env var: OPENAI_API_KEY Supported sizes: 256x256, 512x512, 1024x1024, 1024x1792, 1792x1024

Stability AI

const image = await generateImage({
  provider:        'stability',
  model:           'stable-image-core',   // or 'stable-diffusion-xl-1024-v1-0', 'sd3'
  prompt:          'An art deco travel poster for a moon colony, vintage style.',
  negativePrompt:  'text, watermark, low quality',
  width:           1024,
  height:          1024,
  steps:           30,
  providerOptions: {
    stability: {
      stylePreset: 'illustration',   // 'photographic' | 'illustration' | 'digital-art' | ...
      seed:        42,
      cfgScale:    8,                // classifier-free guidance scale
      sampler:     'K_DPMPP_2M',
    },
  },
});

Default model: stable-image-core Env var: STABILITY_API_KEY Style presets: photographic, illustration, digital-art, anime, comic-book, fantasy-art, line-art, modeling-compound, neon-punk, origami, pixel-art, tile-texture

Replicate

const image = await generateImage({
  provider:    'replicate',
  model:       'black-forest-labs/flux-1.1-pro',
  prompt:      'A product photo of a titanium watch on polished black stone.',
  aspectRatio: '1:1',
  providerOptions: {
    replicate: {
      outputQuality: 90,
      outputFormat:  'webp',
      input: {
        go_fast:          true,
        safety_tolerance: 2,
      },
    },
  },
});

Default model: black-forest-labs/flux-1.1-pro Env var: REPLICATE_API_TOKEN Popular models:

black-forest-labs/flux-schnell — fast, lower quality
black-forest-labs/flux-1.1-pro — high quality
stability-ai/sdxl — SDXL via Replicate
bytedance/sdxl-lightning-4step — 4-step fast generation

OpenRouter

OpenRouter can route image generation requests to multiple backends:

const image = await generateImage({
  provider: 'openrouter',
  model:    'openai/gpt-image-1',   // or 'stability-ai/stable-image-core'
  prompt:   'Abstract geometric art, primary colors, Mondrian style.',
  providerOptions: {
    openrouter: {
      route: 'fallback',   // try next model if primary fails
    },
  },
});

Default model: openai/gpt-image-1 Env var: OPENROUTER_API_KEY

Local Stable Diffusion

Run Stable Diffusion locally with no API costs. Supports both Automatic1111 (A1111) and ComfyUI WebUI APIs.

const image = await generateImage({
  provider: 'stable-diffusion-local',
  model:    'v1-5-pruned-emaonly',   // model checkpoint name
  prompt:   'A brutalist house in dense fog, dramatic lighting.',
  negativePrompt: 'blurry, low quality, text',
  width:    512,
  height:   512,
  steps:    25,
  seed:     1234,
  baseUrl:  'http://localhost:7860',   // or set STABLE_DIFFUSION_LOCAL_BASE_URL
  providerOptions: {
    'stable-diffusion-local': {
      samplerName:    'DPM++ 2M Karras',
      cfgScale:       7,
      restoreFaces:   false,
      hiresUpscaler:  'R-ESRGAN 4x+',
      hiresSteps:     10,
      denoisingStrength: 0.45,
    },
  },
});

Default model: v1-5-pruned-emaonly Env var: STABLE_DIFFUSION_LOCAL_BASE_URL (default: http://localhost:7860)

Provider Options Passthrough

Every provider accepts a providerOptions object keyed by provider ID. Unknown keys are silently ignored — safe to set options for multiple providers:

const image = await generateImage({
  provider: 'stability',
  model:    'stable-image-core',
  prompt:   '...',
  providerOptions: {
    // Only the 'stability' key is used; others are ignored
    stability: { stylePreset: 'photographic' },
    openai:    { style: 'vivid' },    // ignored when provider != 'openai'
    replicate: { outputQuality: 80 }, // ignored when provider != 'replicate'
  },
});

This pattern lets you pre-configure options for all providers and switch with just a single provider field change.

Local Setup (A1111 and ComfyUI)

Automatic1111 WebUI

# Clone and install
git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui
cd stable-diffusion-webui

# Download a model checkpoint (example: SD 1.5)
mkdir -p models/Stable-diffusion
wget -O models/Stable-diffusion/v1-5-pruned-emaonly.safetensors \
  https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors

# Launch with API enabled
./webui.sh --api --listen

# Server starts at http://localhost:7860

Set the env var:

export STABLE_DIFFUSION_LOCAL_BASE_URL=http://localhost:7860

ComfyUI

git clone https://github.com/comfyanonymous/ComfyUI
cd ComfyUI
pip install -r requirements.txt

# Place model checkpoints in models/checkpoints/
# Launch
python main.py --port 7860

export STABLE_DIFFUSION_LOCAL_BASE_URL=http://localhost:7860

AgentOS uses the A1111-compatible /sdapi/v1/txt2img endpoint. ComfyUI requires the A1111 API compatibility layer or a custom provider (see below).

Ollama with Image Model

ollama pull stable-diffusion
export OLLAMA_BASE_URL=http://localhost:11434

const image = await generateImage({
  provider: 'ollama',
  model:    'stable-diffusion',
  prompt:   'A cozy reading nook with warm lamplight.',
});

Custom Image Provider

import {
  generateImage,
  registerImageProviderFactory,
  type IImageProvider,
  type ImageGenerationRequest,
  type ImageGenerationResult,
} from '@framers/agentos';

class MyImageProvider implements IImageProvider {
  readonly providerId    = 'my-provider';
  readonly isInitialized = false;
  readonly defaultModelId = 'my-default-model';

  async initialize(): Promise<void> {
    // Connect to your image service
  }

  async generateImage(request: ImageGenerationRequest): Promise<ImageGenerationResult> {
    const response = await myImageAPI.generate({
      prompt: request.prompt,
      model:  request.modelId,
    });

    return {
      created:    Math.floor(Date.now() / 1000),
      modelId:    request.modelId ?? this.defaultModelId,
      providerId: this.providerId,
      images: [
        {
          url:      response.imageUrl,
          mimeType: 'image/png',
          width:    request.width ?? 1024,
          height:   request.height ?? 1024,
        },
      ],
      usage: { totalImages: 1 },
    };
  }
}

// Register the factory
registerImageProviderFactory('my-provider', () => new MyImageProvider());

// Use it
const image = await generateImage({
  provider: 'my-provider',
  model:    'my-default-model',
  prompt:   'A product photo on white background.',
});

Usage Tracking

Track image generation usage across providers:

import { generateImage, getRecordedAgentOSUsage } from '@framers/agentos';

await generateImage({
  provider: 'openai',
  prompt:   'A banner image for our launch.',
  usageLedger: {
    enabled:   true,
    sessionId: 'launch-campaign',
  },
});

const usage = await getRecordedAgentOSUsage({
  enabled:   true,
  sessionId: 'launch-campaign',
});

console.log(usage.totalImages);     // 1
console.log(usage.estimatedCost);   // e.g., 0.04

HIGH_LEVEL_API.md — full generateImage() API reference
EXAMPLES.md — automated blog publisher example with image generation
GETTING_STARTED.md — installation and environment setup

Table of Contents​

Overview​

generateImage() API​

Common Parameters​

Provider Reference​

OpenAI (gpt-image-1)​

Stability AI​

Replicate​

OpenRouter​

Local Stable Diffusion​

Provider Options Passthrough​

Local Setup (A1111 and ComfyUI)​

Automatic1111 WebUI​

ComfyUI​

Ollama with Image Model​

Custom Image Provider​

Usage Tracking​

Related Guides​

Table of Contents