---
title: "Draft a reply with AI"
method: POST
path: "/support/conversations/{id}/ai-draft"
---

# Draft a reply with AI

`POST /support/conversations/{id}/ai-draft`

## Request

- Base URL: `https://api.web2wave.com/api`
- URL: `https://api.web2wave.com/api/support/conversations/{id}/ai-draft`
- Auth: API key in header `api_key`

Writes an answer from the project's reply templates and the conversation, like "Answer with AI" in the cabinet.
Nothing is saved or sent — pass the text to `…/reply` once you are happy with it. Limited to 20 calls per minute.

## Path parameters

- `id` integer, required

## Headers

- `api_key` string, required
- `X-Web2wave-Project-Id` string

## Request body

- object
  - `draft` string: Text you already started; the AI continues or improves it

## Response `200`

Draft

- object
  - `success` boolean
  - `text` string

## Other responses

- `422`: AI is not configured or the conversation cannot be answered
- `429`: Rate limit
- `502`: AI provider error

## Changes

- **2026-11-03** `eb8b2d9124ac`: 1 info
  - endpoint added

[Operation history](https://skmtc.dev/web2wave/apis/web2wave-api/changes/support/conversations/:id/ai-draft/post.md)

---

[API](https://skmtc.dev/web2wave/apis/web2wave-api.md) - [All operations](https://skmtc.dev/web2wave/apis/web2wave-api/llms.txt) - [OpenAPI document](https://skmtc.dev/web2wave/apis/web2wave-api/revisions/eb8b2d9124ac?raw)