AI

Use this demo with any OpenAI-compatible /v1/chat/completions provider. The endpoint, model, and token fields are used only in this page session and are not stored.

AI providerMode, endpoint, model, token, and the stream, thinking, and reasoning toggles are used only in this page session and are not stored.

Mode

OpenAI-compatible provider

Mode

Endpoint

Endpoint

Model

Model

API token

API token

Stream

Thinking

Show reasoning

Backend endpoint mode

Switch the Mode field to Backend endpoint to let tiptapify talk to your backend directly, without a custom aiProvider. It posts a minimal payload (prompt, instruction, and the optional stream, thinking, reasoning_effort, model fields) to the endpoint and accepts a plain JSON, { content }, or SSE streaming response — see the AI guide for the full contract.

Local LM Studio

For LM Studio, use an endpoint like:

http://localhost:1234/v1/chat/completions

If LM Studio runs on another machine, use that host instead, for example:

http://192.168.0.107:1234/v1/chat/completions

For browser-based demos, use a dev proxy if your provider does not allow CORS.