generated: '2026-07-19'
method: searched
source: https://docs.inference.net/api/async-inference/webhooks/quick-reference
spec_type: Webhooks
docs: https://docs.inference.net/api/async-inference/webhooks/getting-started-with-webhooks
description: >-
  Webhooks deliver results of asynchronous inference (batch and grouped requests).
  Each delivery is a JSON envelope with top-level `event`, `timestamp`, and
  `webhook_id`, plus a `data` object holding the result (state, message, original
  request, and an OpenAI-format `response`). Individual events carry `generation_id`;
  batch/group events carry `group_id`.
events:
- name: generation.completed
  description: Fires when a chat completion finishes.
- name: async-embedding.completed
  description: Fires when an embedding request completes.
- name: slow-group.completed
  description: Fires when a batch of grouped requests finishes.
headers:
- X-Inference-Event
- X-Inference-Webhook-ID
- X-Inference-Generation-ID
- X-Inference-Group-ID
- User-Agent
- Content-Type
envelope:
  fields:
  - event
  - timestamp
  - webhook_id
  - data
  data_fields:
  - state
  - message
  - request
  - response
notes:
- The quick-reference did not document a signature/HMAC verification header; the
  listed X-Inference-* headers are informational. Not asserting signed webhooks.