# Image Generation ## OpenAPI Specification ```yaml openapi: 3.0.1 info: title: '' description: '' version: 1.0.0 paths: /v1/aiart/imagegeneration/{provider}: post: summary: Image Generation deprecated: false description: >+ The AI Image Generation API allows you to generate images based on a combination of text prompts and reference images. It supports multi-modal input, enabling scenarios such as character consistency, style transfer, composition guidance, and creative image generation. With this API, developers can: Generate new images purely from textual descriptions (e.g., landscapes, portraits, product shots) Provide reference images to maintain character or object consistency Combine multiple reference images with prompts for more precise control Customize the output resolution via width and height parameters Use smart reference mode (smart_reference) to automatically extract key visual features from the reference image. ## Gemini 2.5 Flash Image | Aspect Ratio | Resolution | Tokens | |--------------|------------|--------| | 1:1 | 1024×1024 | 5 | | 2:3 | 832×1248 | 5 | | 3:2 | 1248×832 | 5 | | 3:4 | 864×1184 | 5 | | 4:3 | 1184×864 | 5 | | 4:5 | 896×1152 | 5 | | 5:4 | 1152×896 | 5 | | 9:16 | 768×1344 | 5 | | 16:9 | 1344×768 | 5 | | 21:9 | 1536×672 | 5 | ## Gemini 3 Pro Image (Preview) | Aspect Ratio | 1K Resolution | 1K Tokens | 2K Resolution | 2K Tokens | 4K Resolution | 4K Tokens | |--------------|----------------|-----------|----------------|-----------|----------------|-----------| | 1:1 | 1024×1024 | 15 | 2048×2048 | 15 | 4096×4096 | 30 | | 2:3 | 848×1264 | 15 | 1696×2528 | 15 | 3392×5056 | 30 | | 3:2 | 1264×848 | 15 | 2528×1696 | 15 | 5056×3392 | 30 | | 3:4 | 896×1200 | 15 | 1792×2400 | 15 | 3584×4800 | 30 | | 4:3 | 1200×896 | 15 | 2400×1792 | 15 | 4800×3584 | 30 | | 4:5 | 928×1152 | 15 | 1856×2304 | 15 | 3712×4608 | 30 | | 5:4 | 1152×928 | 15 | 2304×1856 | 15 | 4608×3712 | 30 | | 9:16 | 768×1376 | 15 | 1536×2752 | 15 | 3072×5504 | 30 | | 16:9 | 1376×768 | 15 | 2752×1536 | 15 | 5504×3072 | 30 | | 21:9 | 1584×672 | 15 | 3168×1344 | 15 | 6336×2688 | 30 | ## Seedream 4.0 | Resolution | Tokens | |--------------|------------| | Total pixels range: [1280×720, 4096×4096] | 5 ## Qwen | Resolution | Tokens | |--------------|------------| | Total pixels range: [1280×720, 2048×2048] | 3 | tags: - Reference parameters: - name: provider in: path description: > Specifies the AI model used for image generation. This field is an enumeration that defines which backend model will process the request. Different models provide different capabilities, image quality levels, aspect-ratio support, performance characteristics, and token costs. ## Allowed Values (Enum) | Enum Value | Model Description | | ------------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **qwen** | Great at reading and creating clear text inside images. | | **seedream-4-0-250828** | Precise control over multiple images. | | **gemini-2.5-flash-image** | Unparalleled facial consistency. | | **gemini-3-pro-image-preview** | Ultra-sharp 4K, consistent, intelligent generation. | required: true example: gemini-3-pro-image-preview schema: type: string - name: Authorization in: header description: >- Here, is the API key you obtained from the Fotor Developer Platform. required: true example: Bearer schema: type: string requestBody: content: application/json: schema: type: object properties: content: type: array items: type: object properties: type: type: string description: >- The type of input. Allowed values: text (text prompt), image_url (reference image). text: type: string description: >- Used when type = "text". Describes the desired content, style, composition, etc. url: type: string description: >- Used when type = "image_url". A URL pointing to the accessible reference image. role: type: string description: >- Indicates the role/type of the reference image, guiding the model to focus on certain aspects of the image. Common value: smart_reference (intelligent reference). required: - type x-apifox-orders: - type - text - url - role description: >- Describes the input elements for image generation. Supports mixed text prompts and reference images. width: type: integer description: Target width of the generated image (in pixels). height: type: integer description: Target height of the generated image (in pixels). required: - content - width - height x-apifox-orders: - content - width - height example: content: - type: text text: >- Subject: The human subject is reimagined as the main character in a cinematic moment, walking a loyal pet companion. Their silhouette is enhanced with soft, dramatic lighting that adds depth, presence, and storytelling. The subject’s features remain natural, recognized, and elevated with movie-grade atmosphere. Outfit: A stylish, film-inspired outfit adapted to the user’s daily-wear vibe — flowing coats, layered sweaters, subtle textures, or minimalist modern clothing depending on the final look. Everything is realistic, grounded, and aesthetically pleasing. Pet: A [Dog] walking beside the subject. The pet’s form is natural, expressive, and receives soft directional lighting to match the cinematic tone. The pet behaves calmly, walking close to the subject. Setting: A beautifully cinematic environment such as a quiet evening street, a misty park pathway, a soft golden-hour field, or a lantern-lit city walkway. Warm or cool tones depending on the mood. Light mist, drifting particles, or soft bokeh enhance the storytelling. Nature or city elements glow gently in the background. Camera / Composition: 4:5 vertical movie-frame Medium-distance shot showing the full walk Cinematic depth, wide aperture feel, atmospheric haze No text, no watermark - type: image_url url: >- https://u-static.fotor.com/uid_0f5eccd484e6403084043fd20e320027/image_to_image/app-fotor-web/ef3f9d2c-aa9b-4a99-96e6-c6c32d11df36.jpg role: smart_reference width: 841 height: 1152 responses: '200': description: '' content: application/json: schema: type: object properties: code: type: string description: '000: Success，999: Unknown error' x-apifox-mock: '000' examples: - '000' msg: type: string x-apifox-mock: success examples: - success data: type: object properties: taskId: type: string description: Task ID, used to query task information. x-apifox-mock: e0dc194703174b6eafeacb4ee8e02850 examples: - e0dc194703174b6eafeacb4ee8e02850 required: - taskId x-apifox-orders: - taskId required: - code - msg - data x-apifox-orders: - code - msg - data headers: {} x-apifox-name: success security: [] x-apifox-folder: Reference x-apifox-status: released x-run-in-apifox: https://app.apifox.com/web/project/5562290/apis/api-384299001-run components: schemas: {} securitySchemes: {} servers: - url: https://api-b.fotor.com description: 正式环境 security: [] ```