Sign in
← AI PulseAug 20, 2026

Wire · news · Single-source brief

xAI Introduces Deferred Chat Completions for Long-Running Inference

xAI has launched Deferred Chat Completions, enabling users to initiate a chat completion request and retrieve the result later, within a 24-hour window.

By Illumora Editorial

Source · Aug 20, 2026, 3:13 PM · On Illumora · Aug 20, 2026, 3:38 PM

Media from the primary source — shown here so you can stay on Illumora.

Rewritten from one allowlisted primary — not independent enterprise reporting. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →xAI Docs — Deferred Chat Completions | SpaceXAI Docs
Save

xAI has introduced Deferred Chat Completions, a new feature designed for handling long-running inference requests. This functionality is accessible through REST requests or the xAI SDK.

Deferred Chat Completions allow users to submit a request, receive a response_id, and then retrieve the full completion result at a subsequent time. The result remains available for retrieval exactly once within a 24-hour period before it is discarded.

Key Points

  • Deferred Chat Completions are designed for long-running inference requests.
  • They are available via REST requests or the xAI SDK.
  • Users receive a response_id after initiating a deferred completion.
  • The completion result can be retrieved exactly once within 24 hours.
  • After 24 hours, the result is discarded.
  • The deferred completion rate limit matches the standard chat completions rate limit.
  • When a completion result is not ready, the request returns a 202 Accepted status with an empty response body.
  • The model's raw thinking trace is accessible via message.reasoning_content in the chat completion response.

Context

According to xAI, this feature addresses the need for asynchronous processing of chat completions, particularly for tasks that may require extended inference times. Users can monitor their rate limits for deferred completions through the xAI Console.

Why It Matters

This capability provides developers with a mechanism to manage potentially lengthy AI inference tasks without requiring an immediate, synchronous response. It allows for more flexible integration of xAI's models into applications that can benefit from asynchronous processing, such as background tasks or workflows with variable completion times.

What To Do

  • Review the xAI documentation for Deferred Chat Completions to understand implementation details.
  • Test the Deferred Chat Completions feature using either REST requests or the xAI SDK for long-running inference tasks.
  • Compare the rate limits for deferred completions with standard chat completions in the xAI Console.
  • Implement retry logic for retrieving results, handling 202 Accepted responses when the completion is not yet ready.

Keep Exploring

/atlas/grok-family