Sign in
← AI PulseAug 20, 2026

Wire · news · Multi-source brief

xAI Introduces Speech to Speech API for Real-Time Voice Conversations

xAI has launched a Speech to Speech API that facilitates real-time voice conversations over WebSocket, with pricing based on audio duration and text input messages.

By Illumora Editorial

Source · Aug 20, 2026, 3:13 PM · On Illumora · Aug 20, 2026, 4:04 PM

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

Synthesized from multiple allowlisted primaries on the same event. 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 — Speech to Speech | SpaceXAI Docs
Save

xAI has released a new Speech to Speech API designed for real-time voice conversations, operating via WebSocket. This API is billed by the minute of audio and includes a flat fee for each text input message. The service supports function calling, integrating capabilities such as web search, X search, collections, MCP, and custom functions.

Key Points

  • The Speech to Speech API enables real-time voice conversations over WebSocket.
  • Audio pricing for the API starts at $0.05 per minute or $3.00 per hour.
  • Text input messages are billed at $0.004 per message for each conversation.item.create event.
  • The API supports function calling, including web search, X search, collections, MCP, and custom functions.
  • The service operates in the us-east-1 region.
  • grok-4.6 is mentioned in the context of the Speech to Speech API and Collections service.

Context

According to xAI documentation, the Speech to Speech API charges are based on the duration of audio sent or received, and for text events that are sent without accompanying audio. Specifically, each conversation.item.create event is billed, with exceptions for response.create events and function_call_output items, which are not billable. Items containing input_audio or audio are billed under the audio meter. The Collections service, which can be integrated via function calling, allows users to upload and search documents, supporting various MIME types and offering features like semantic search and metadata fields, according to xAI. grok-4.6 is described by xAI as a frontier model for coding, agentic tasks, and knowledge work, featuring reasoning capabilities.

Why It Matters

This API introduces a direct method for developers to integrate real-time voice interactions into applications, potentially altering how users engage with AI systems. The pricing structure provides clarity on the costs associated with both audio and text components of these conversations, which is crucial for builders managing operational expenses.

What To Do

  • Open the xAI documentation to review the Speech to Speech API guide for getting started with real-time voice conversations.
  • Compare the audio and text input pricing tiers to estimate potential costs for planned applications.
  • Test the Speech to Speech API in the Playground to understand its real-time conversational capabilities.
  • Note the supported function calling features, such as web search and collections, for potential integration into applications.

Keep Exploring

/atlas/**grok**-family