> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://developer.ideogram.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://developer.ideogram.ai/_mcp/server.

# Generate Images

> Choose the right Ideogram image generation endpoint — Ideogram 4.0, P-Image, or Ideogram 3.0, with synchronous, asynchronous, and transparent-background variants.

These endpoints create new images from a text prompt. Pick the model family first, then the delivery mode.

## Which model should I use?

| Model                | Best for                             | Endpoint                                  |
| -------------------- | ------------------------------------ | ----------------------------------------- |
| **Ideogram 4.0**     | Highest quality, best text rendering | `POST /v1/ideogram-v4/generate`           |
| **P-Image Ideogram** | Low-latency generation               | `POST /v1/text-to-image/p-image-ideogram` |
| **Ideogram 3.0**     | Previous generation, stable contract | `POST /v1/ideogram-v3/generate`           |

## Delivery modes

* **Synchronous** endpoints return the finished images in the response.
* **Asynchronous** endpoints (under `/async/`) return a `generation_id` immediately. Retrieve results by polling `GET /v1/generations/{generation_id}` or by supplying a `webhook_url` (see the Webhooks guide).
* **Transparent background** variants return images with alpha channels, suited for logos, stickers, and compositing.