TinyMCE AI API Overview

The TinyMCE AI REST API enables AI integration within your application using the same infrastructure as the TinyMCE AI plugin, but without being tied to the editor. Integrate with multiple AI agents via one API, accessing all functionality from the Chat, Review, and Quick Actions features, with your own custom UI and workflows.

API Reference: The complete API documentation with all endpoints, parameters, request/response schemas, and interactive examples is available at TinyMCE AI API Documentation

Getting started

New to TinyMCE AI? Start with the Quick Start guide to set up the environment, generate access credentials, and make the first API call.

TinyMCE AI features

  • Chat: Interactive AI chats with history and persistent context.

  • Review: Content analysis and proofreading, optimized for larger content.

  • Quick Actions: Fast, stateless operations for specific tasks.

Architecture

The following pages cover the AI service architecture and configuration.

  • Models: AI model selection, capabilities, and configuration.

  • Streaming: Real-time streaming of AI-generated responses.

  • Permissions: How to control user access to features.

  • Limits: Rate limits, context size limits, and file restrictions.

Versioning

The version of each endpoint is part of its path, for example /v1/conversations. The TinyMCE AI plugin calls the /v1 endpoints, and requests the model list with compatibility version 1. The compatibility version is separate from the endpoint version. See Model compatibility versions.

Pagination

List endpoints use cursor pagination. These include the conversation list, and the documents, files, and web resources of a conversation. The API reference documents the pagination parameters on each list operation, for example Fetch multiple conversations.

Error responses

Errors returned by the AI service use a JSON body with these fields:

Field Description

statusCode

The HTTP status code.

code

A machine-readable error code, for example missing-permissions.

message

A short description of the error.

traceId

A unique identifier for the request. Include it when contacting Tiny Support.

data

Optional. An object with details specific to the error.

explanation

Optional. More detail about why the error occurred.

action

Optional. A suggested next step.

For example, a token without the permission a request needs returns:

{
  "statusCode": 403,
  "code": "missing-permissions",
  "message": "The user is missing permissions",
  "traceId": "8f3c1c2e-5b7a-4d0e-9a61-2f4c8e0b7d13",
  "data": {
    "missingPermissions": ["ai:conversations:write"]
  }
}

Error codes

The API reference lists every code in its Error codes section. For token errors, see JWT troubleshooting. For a missing-permissions response, see Permissions.

Rate limit errors

For 429 responses with the code rate-limits-exceeded, see Handling rate limit errors.

Resources and support