Text to image
Submit a prompt, model, size, and aspect ratio for covers, product imagery, marketing assets, illustration, and batch content.
POST /images/generationsAI VIDEO GENERATION API
One API for leading generation models
Call multiple image and video generation models through one server-side API and bring them into products and automated workflows quickly.
Built for developers familiar with HTTP, JSON, and server-side secret management. The live model catalog is the source of truth.
One task protocol across multiple models
- GPT Image 2
- Nano Banana 2
- Nano Banana Pro
- Seedream 5.0
- Qwen Image 2.0
- Midjourney V7 Imagine
- Seedance 2.0
- Grok Video 3
Generation capabilities
From still images to motion video
Provider differences stay out of your business layer. Choose a model, submit parameters, then read the result through one task lifecycle.
Image generation and editing
Use public HTTPS reference images for background replacement, style transfer, local edits, multi-image reference, or character-consistency workflows.
mode: editText and image to video
Depending on the model, submit text, a first frame, reference video, or audio; then retrieve landscape or portrait video by task ID.
POST /videos/generationsA real request
GPT Image 2 uses the same task API
For gpt image2 api or GPT-Image-2 API integrations, the Flux Art model ID is gpt-image-2. It is not a drop-in OpenAI SDK replacement, but it shares authentication, idempotency, and task queries with the other image and video models.
- Auth
Authorization: Bearer fa_live_...- Idempotency
- Use a new
Idempotency-Keyfor every new task - Response
- Receive a task ID first, then read the image URL when complete
curl -X POST \
'https://open-api.flux-art.ai/openapi/v1/images/generations' \
-H 'Authorization: Bearer <API_KEY>' \
-H 'Idempotency-Key: 4e4be4c5-4c59-48ef-893b-f9419af9f8e8' \
-H 'Content-Type: application/json' \
-d '{
"model": "gpt-image-2",
"mode": "generate",
"prompt": "A clean product photograph of a ceramic lamp",
"size": "1K",
"aspect_ratio": "16:9"
}'
# 201 Created
{
"data": {
"id": d2b22844-...",
"status": queued",
"model": gpt-image-2"
}
}Integration flow
Complete a server integration in three steps
The first release uses asynchronous tasks and polling, making it suitable for web services, background jobs, scripts, and automation platforms.
Create and store an API Key
The full secret is shown once. Put it in a server-side secret manager or environment variable—never browser code, Git, or logs.
Submit a generation task
Choose the model and parameters from
/openapi/v1/models, generate a unique idempotency key, then call the image or video endpoint.Poll and store the result
Save the task ID and poll at the recommended interval. When it succeeds, read the image or video URL and copy it into your own storage.
Account and task rules
The API follows the same business rules as the web app
OpenAPI reuses Flux Art model validation, dynamic pricing, provider routing, point deductions, and failure refunds instead of maintaining a second hidden rule set.
- Membership, points, and prices stay aligned
- The API has no separate quota and cannot bypass video membership requirements or free-user daily image limits.
- Web and API tasks share concurrency
- Web tasks consume API concurrency and vice versa. Requests over the concurrency limit are not charged.
- Failures use idempotent refunds
- Final provider failures follow the existing refund rules; verify them through
usage.points_refunded. - API queries are isolated from web history
- An API Key can only query OpenAPI tasks for the current account and cannot read private prompt history from the web app.
Use cases
For products that automate content generation
For occasional manual generation, the Flux Art web app is simpler. Use the API when generation becomes part of a business workflow.
Content creation tools
Let users create covers, illustrations, product images, ads, or short videos in your product while you control the interaction and workflow.
Batch workflows
Connect generation to a CMS, catalog, social schedule, or internal operations script and reduce repetitive browser work.
System integration
Use a shared model catalog, task lifecycle, and error handling in a website, app backend, bot, or automation platform.
FAQ
Flux Art API frequently asked questions
Straightforward answers about supported models, accounts, billing, security, and getting started.
Flux Art API is a model access service for products and automated workflows. One API key can call supported AI image and video models without maintaining separate integrations for every provider.
It is for teams that want AI image or video generation inside a website, app, internal tool, or content workflow. Common uses include ecommerce product images, marketing assets, batch content generation, and video creation.
Yes. Flux Art API supports GPT Image 2, Nano Banana, Grok, Seedance, and other image and video models. Check the API reference and live model catalog for the current model IDs and parameters.
No. Flux Art manages the supported model integrations. You only need your Flux Art account and API key.
Yes. One active API key can call the image and video models available to your account, so you do not need a separate key for every model.
Both use the same account, credit balance, and model capabilities. The web app is convenient for hands-on creation, while the API is designed for product integration, batch generation, and automated workflows.
API key creation is currently available to paid members. Free users who choose to create a key will see the available upgrade plans and can manage a key from the same account after upgrading.
The API shares account credits, membership benefits, and promotional discounts with the web app. Usage depends on the model and generation settings and is reported in the task response. Final failures that qualify under the refund rules return the corresponding credits.
Generation time varies by model, task type, output settings, and queue conditions. A successful submission returns a task ID immediately, so your service can continue other work while checking the task status.
Yes, when the selected model supports reference images. Reference media must use a public HTTPS URL that the model service can access, and each model may have different format, size, and image-count limits.
Store it only in server-side environment variables or a dedicated secret manager. Never expose it in browser code, public repositories, app bundles, logs, or shared chat messages.
Usually not. Flux Art API uses shared authentication and task handling across supported models. You mainly change the model ID and model-specific parameters such as size, resolution, or reference inputs.
Generated content can be used for ecommerce, advertising, social media, and other commercial projects, subject to the Flux Art terms, applicable model rules, and third-party copyright, trademark, and likeness rights.
Sign in, upgrade to an eligible plan, create an API key, and open the API reference. Start by checking the live model catalog, submit one image task from a code example, and then query the returned task ID for the result.
Confirm model capabilities, then integrate
After creating a key, validate parameters from the model catalog and run one real task through submission, polling, result storage, and failure retries.