Skip to main content
DimiLinks provides a unified API that follows familiar OpenAI and Anthropic conventions. A single Bearer key lets you call text models including Claude and GPT, submit gpt-image-2 image tasks, and submit grok-imagine-video video tasks.
We recommend using the asynchronous workflow for both image and video generation: submit a task, then poll the task endpoint for the result.

Quickstart

Use one curl command to verify authentication and make your first chat request.

Chat completions

Call GPT, Claude, and other text models through the OpenAI-compatible protocol.

Anthropic Messages

Use the native Anthropic protocol to call the full Claude model family.

Text to image / gpt-image-2

Generate and edit images at 1K, 2K, or 4K, with asynchronous task retrieval.

Text to video / grok-imagine-video

Generate 6- to 20-second videos from text or reference images.

IDE / CLI integrations

Connect Claude Code, Codex CLI, Cursor, Continue, Cherry Studio, and more.

​
Request URL

All DimiLinks model APIs use this request URL:
Use this value directly as the base URL for OpenAI-compatible APIs. Native clients such as the official Anthropic SDK and Claude Code append /v1/messages automatically, so configure their base_url as https://api-direct.dimilinks.com/ to avoid an invalid /v1/v1/messages request. Both forms use the same domain and Bearer key.

​
Common destinations