---
title: "One-shot synchronous ask - text in, the agent's reply out (b2b)"
method: POST
path: "/v2/agents/{id}/reply"
tags: ["Agents"]
---

# One-shot synchronous ask - text in, the agent's reply out (b2b)

`POST /v2/agents/{id}/reply`

## Request

- Base URL: `https://api.chat.ethora.com`
- URL: `https://api.chat.ethora.com/v2/agents/{id}/reply`
- Auth: HTTP bearer

Sends the text straight to the agent's LLM persona (plus its RAG
knowledge base when enabled) and returns the reply in the same HTTP
response - no chat room or XMPP involved. Built for server-to-server
automations (e.g. n8n flows). Requires a b2b token: the call bypasses
room membership entirely. Responses can take 10-30s for long
generations - set client timeouts accordingly.

## Path parameters

- `id` string, required

## Request body

- object
  - `text` string, required

## Response `200`

ok ({ ok, agentId, reply, model, ragDocsUsed, totalTokens })

## Other responses

- `400`: Bad Request
- `401`: Unauthorized
- `403`: Forbidden
- `404`: agent not found
- `422`: validation error
- `500`: Internal Server Error
- `503`: AI service not configured

---

[API](https://skmtc.dev/ethora/apis/platform-api.md) · [All operations](https://skmtc.dev/ethora/apis/platform-api/llms.txt) · [OpenAPI document](https://skmtc.dev/ethora/apis/platform-api/revisions/01b6e64378ee?raw)