i have the conversational ai agent read a long script i think like for 3 minutes. but after that it | ElevenLabs | Page 1

true ospreyBOT Feb 23, 2026, 5:45 PM

#

AI-generated Answer for mattyg1997

Likely cause: your agent’s Turn timeout is too short, so immediately after it finishes the 3‑minute read it hits the “no‑input” prompt (“didn’t catch that”).

What to do

Increase Turn timeout: In the agent’s Advanced tab → Turn Timeout, raise the duration (allowed range is 1–30 seconds). For long monologues/disclaimers, use a longer value (e.g., 15–30s) so callers have time to respond before the reprompt triggers. (elevenlabs.io)
Consider interruptions: If you want callers to be able to speak over the script (barge‑in), enable Interruptions in Advanced → Client Events. If you need the script read in full (e.g., legal text), disable interruptions and rely on the longer Turn timeout. The docs explicitly recommend longer timeouts and disabling interruptions for legal disclaimers. (elevenlabs.io)

Reference

Conversation flow (Turn timeout, Interruptions, recommended configurations). (elevenlabs.io)

Sources:
Conversation flow | ElevenLabs Documentation
Conversation flow | ElevenLabs Documentation
Conversation flow | ElevenLabs Documentation

fast ibex Feb 23, 2026, 5:59 PM

#

not 4 minute ? will 30 seconds work for like a 4 minute worth of script

true ospreyBOT Feb 23, 2026, 5:59 PM

#

AI-generated Answer for mattyg1997

No—turn timeout can’t be set to 4 minutes. ElevenAgents supports a turn-timeout range of 1–30 seconds only, configured in the agent’s Advanced settings. The timeout controls how long the agent waits in user silence after it finishes speaking, and it’s independent of how long the script was. So the maximum you can set is 30 seconds. For long monologues (e.g., disclaimers), you can disable interruptions to ensure the agent finishes speaking, but the post-speech wait limit still caps at 30 seconds. See Conversation flow → Turn timeout. https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow

fast ibex Feb 23, 2026, 6:01 PM

#

but i actually want barge in

true ospreyBOT Feb 23, 2026, 6:01 PM

#

AI-generated Answer for mattyg1997

You can enable barge-in (interruptions) for ElevenLabs agents.

Dashboard: In your agent, go to Advanced → Client Events and enable Interruptions. This allows users to interrupt while the agent is speaking. (elevenlabs.io)
SDK/Web: Make sure the agent has the Interruptions client event enabled; otherwise callbacks won’t fire in the SDK. (elevenlabs.io)
Telephony:
- Genesys: Only Audio and Interruption events are supported—ensure Interruption events are enabled. (elevenlabs.io)
- Vonage WebSocket connector: The connector handles conversation events including interruption. (elevenlabs.io)
Programmatic barge-in (advanced): If you’re generating long responses, use the Multi-Context WebSocket to handle user barge-ins by closing the current speech context and starting a new one when the user speaks; you can also stream in chunks and flush at sentence boundaries. (elevenlabs.io)

If you tell me your integration path (web widget/SDK vs. Genesys/Vonage), I can point you to the exact steps. (elevenlabs.io)

fast ibex Feb 23, 2026, 6:03 PM

#

but i still have that problem of a 4 minute read out script then after that it automatically ask sorry i didnt catch that

#i have the conversational ai agent read a long script i think like for 3 minutes. but after that it