Inworld AI · Rate Limits
Inworld Ai Rate Limits
Rate-limit policies for the Inworld AI platform. The Realtime API caps concurrent sessions per workspace and per-session packet rate. TTS, STT, and Router endpoints share an HTTP request-per-minute (RPM) bucket with tier-based scaling. Synthesis is additionally capped at 2,000 input characters per request and 16 MB of output audio per response.
Inworld Ai Rate Limits is the machine-readable rate-limit profile for Inworld AI on the APIs.io network, conforming to the API Commons Rate Limits specification.
It captures 4 rate-limit definitions.
The profile also includes response codes documented for throttled and quotaExceeded.
Tagged areas include AI, Rate Limiting, Voice, and Realtime.
4 Limits
Throttle: 429
Quota: 429
AIRate LimitingVoiceRealtime
Limits
workspace
request
Per-request synthesis caps. Tier-based RPM applies across the workspace.
request
Configurable end-of-turn detection plus per-request inactivity timeout. Per-tier RPM applies across the workspace.
workspace
Upstream model provider rate limits apply on top of the Inworld tier RPM bucket. The Router returns the upstream 429 response when an upstream provider throttles.