canonical: https://jentic.com/apis/all-images.ai/all-images-ai

# All-Images.ai API

Jentic publishes the only available OpenAPI specification for All-Images.ai API, keeping it validated and agent-ready. All-Images.ai is an AI image generation platform that exposes a tight surface area: one endpoint to kick off a generation job and one endpoint to poll for the resulting image. Authentication is via an X-API-Key header, and jobs run asynchronously so callers submit a prompt and check back for the rendered output. The minimal endpoint count makes it well suited for embedding inside agent workflows where one specific image is needed.

## For AI agents

Generate AI images from natural-language prompts and retrieve them by ID - a two-call flow that fits any agent loop.

## Scope

Does not handle video generation, audio synthesis, or 3D rendering - use for text-to-image generation and retrieval only.

## Capabilities

- Submit a text prompt to generate a new AI image
- Poll for an image by its imageId until generation completes
- Retrieve metadata and the rendered URL for a finished image
- Authenticate every call with a single X-API-Key header
- Drive image creation from agent prompts without browser automation

## Use cases

### Automated marketing creative

Marketing teams generate campaign visuals on demand by sending prompts derived from product copy. The agent posts to `/v1/images/generate`, polls `/v1/images/{imageId}` until ready, and writes the result URL into a CMS record. The flow is two endpoints, so the integration fits inside a single workflow step rather than a multi-stage pipeline.

Example prompt: POST `/v1/images/generate` with prompt 'minimalist product hero on white background', then poll GET `/v1/images/{imageId}` every 5 seconds until status is complete and store the URL.

### Programmatic blog illustration

Editorial bots generate a header image for each new article by feeding the article title into the generation endpoint. Because the API uses a job-and-poll model, the bot can fire off all generations in parallel and reconcile results as they arrive. This keeps publishing pipelines unblocked even when render time spikes.

Example prompt: Generate a header image for the article titled 'How to ship faster' by calling POST `/v1/images/generate`, then poll until the image is ready.

### Agent-driven creative iteration

An agent receives a brief, generates an image, evaluates the result against the brief using a vision model, then re-prompts if the result misses. All-Images.ai's two-endpoint surface is small enough that the loop runs cleanly. Through Jentic the agent never sees the raw API key - it just searches, loads, and executes.

Example prompt: Use Jentic to search 'generate an AI image', execute POST `/v1/images/generate` with the brief, fetch the result, and if a vision check fails, re-prompt with adjusted text.

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | `/v1/images/generate` | Submit a prompt and start image generation |
| GET | `/v1/images/{imageId}` | Get image details and rendered URL |

## Key resources

- **Image generation** — Submit a prompt and start an asynchronous render
- **Image retrieval** — Look up an image by ID and check completion status

## Why Jentic

- **Setup:** Wiring the All-Images.ai API by hand means attaching its API key to api.all-images.ai, submitting generation requests, and polling for the finished image yourself. Through Jentic you install once, import the All-Images.ai API from the API Directory, store the API key once, and your agent calls it.
- **Permission scoping:** The All-Images.ai API puts the image id in the URL path (`/v1/images/{imageId}`) for retrieval while generation targets travel in the request body, so limit the agent to the operations it needs, such as generating an image or fetching one by id. You choose the operations it may call, so generation is not included unless you add it.
- **Credential handling:** Your All-Images.ai API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'generate an image from a prompt' or 'retrieve a generated image', and Jentic returns the matching All-Images.ai operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

## Related APIs

- **OpenAI** — OpenAI's image generation endpoints cover similar text-to-image workflows
- **Stability AI** — Stability AI offers Stable Diffusion-based generation with model selection
- **Replicate** — Replicate hosts many open-source image models behind a unified API
- **remove.bg** — Strip backgrounds from generated images for product use

## FAQ

### Why is there no official OpenAPI spec for All-Images.ai API?

All-Images.ai does not publish an OpenAPI specification. Jentic generates and maintains this spec so that AI agents and developers can call All-Images.ai API via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.

### What authentication does the All-Images.ai API use?

Every call requires an X-API-Key header. Through Jentic, the key is stored in the encrypted vault and injected at execution time, so it never appears in agent prompts or logs.

### Can I generate an image and get the URL in a single call?

No. Generation is asynchronous: POST `/v1/images/generate` returns an imageId, and you poll GET `/v1/images/{imageId}` until the rendered URL is available. Plan for retry logic in the calling code.

### How do I generate an image via the All-Images.ai API through Jentic?

Run pip install jentic, then in code search for 'generate an AI image', load the input schema for POST `/v1/images/generate`, execute with your prompt, and poll the returned imageId via GET `/v1/images/{imageId}` until complete.

### What are the rate limits for the All-Images.ai API?

The OpenAPI spec does not declare explicit rate limits. Treat the asynchronous job model as the throttle and design polling intervals (e.g. 5 seconds) to avoid spamming the retrieval endpoint.

### Can I retrieve an image after the original job has finished?

Yes. GET `/v1/images/{imageId}` returns the image record including the rendered URL as long as the imageId is known. Store the imageId from the generation response if you need to fetch the image later.

### Can I limit what my agent is allowed to do with the All-Images.ai API?

Yes. Because you run Jentic One yourself, your own rules decide which of the two All-Images.ai operations the agent may call, so you can grant it only POST `/v1/images/generate` to start a render, only GET `/v1/images/{imageId}` to fetch a finished image by id, or both. You also control which stored credential the agent uses, so a read-only agent can retrieve images without ever being able to start new generation jobs. Nothing is callable unless you explicitly allow it.