Modes
By the end of this page you will know which interaction modes your agent supports and how to enable them.
Prerequisites
- An AI Agent open in the editor (Create an AI Agent).
- Admin or Developer role.
The four modes
A mode is how your agent is used. qRaptor supports four:
| Mode | Description | Where it’s configured |
|---|---|---|
| Conversation | Chat sessions with memory, via text messages. | Editor (Modes tab) |
| Voice Starter | Real-time voice using STT/TTS or a native realtime model. | Editor (Modes tab) |
| Scheduled | Cron-triggered autonomous runs. | On a deployment |
| API | Stateless REST endpoint for integrations. | On a deployment |
Enable Conversation and Voice
The Modes tab shows the two modes you configure in the editor. At least one mode must stay enabled.
Toggle a mode on
Enable Conversation for text chat, and/or Voice for real-time voice. Turning on Voice reveals the voice configuration panel.
Voice requires a Starter (or higher) plan. When enabled, configure it on the Voice configuration page — TTS model and voice, pipeline mode, call behavior, and the voice persona. To take calls from real phones, the agent also needs a telephony number.
Scheduled and API modes
Scheduled and API modes aren’t toggled in the editor. They become available once the agent is deployed — either added to an existing deployment or deployed on its own.
- Scheduled — runs the agent on a cron schedule with a fixed goal and inputs. Use it for recurring autonomous jobs (daily digests, periodic checks).
- API — exposes the agent as a stateless REST endpoint for integrations, with its own rate limit and API-key requirement.
To use Scheduled or API mode, deploy the agent first. See Deploy.
Save your changes
Click Save in the header to persist your mode selection into a draft version.
Common issues
- Can’t turn off the last mode — At least one mode must remain enabled. Enable another mode first.
- Voice is unavailable — Voice requires a Starter or higher plan.
- No Scheduled/API options in the editor — These are configured on a deployment, not in the editor. Deploy the agent first.