xAI Docs

xAI Docs release notes (June 24, 2026)

Rate limits

• Clarified that limits are enforced on requests per second (RPS) and tokens per minute (TPM). RPS values are derived from the per-minute budget to prevent burst traffic.
• Updated the per-model limits table to display RPS instead of RPM, with corresponding lower numeric values.
• Replaced previous high-RPM entries for grok-imagine-image, grok-imagine-image-quality, grok-imagine-video, and grok-imagine-video-1.5 with the current lower RPS values (and TPM shown as “—”).

Text-to-speech

• Removed the “Requests per minute” row from the limits table, leaving only the RPS value.

Speech-to-text

• Removed the RPM row from the limits table, leaving only the RPS values for both REST and streaming.