bolna_create_agent
Create a voice AI agent with custom system prompt, voice, LLM, and call settings for automated phone conversations.
Instructions
Create a new Bolna voice AI agent with full configuration and system prompt
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| gender | No | Gender of the agent: "female" or "male". If not set, it is auto-detected. | |
| language | No | Transcription language code, e.g. "en", "hi", "ta", "te" | hi |
| voice_id | No | Voice ID from bolna_list_voices, e.g. "vYENaCJHl4vFKNDYPr8y" (Riya Rao Indian Female) | |
| llm_model | No | LLM model, e.g. "gpt-4o-mini", "gpt-4o", "claude-3-5-haiku" | gpt-4o-mini |
| agent_name | Yes | Unique name for the agent | |
| agent_type | No | Agent type, e.g. "other" | other |
| voice_name | No | Voice name, e.g. "Riya Rao - Professional Voice" | |
| voice_model | No | TTS model, e.g. "eleven_turbo_v2_5", "eleven_turbo_v2", "bulbul:v3" | eleven_turbo_v2_5 |
| webhook_url | No | Webhook URL to receive call events | |
| system_prompt | Yes | LLM system prompt that governs agent behaviour | |
| call_terminate | No | Maximum call duration in seconds | |
| voice_provider | No | TTS provider for the voice | elevenlabs |
| incremental_delay | No | Delay in ms between incremental responses | |
| telephony_provider | No | Telephony provider, e.g. "plivo" (default), "exotel" | plivo |
| ambient_noise_track | No | Ambient noise track, e.g. "coffee-shop" | |
| hangup_after_silence | No | Seconds of silence before hanging up | |
| agent_welcome_message | No | First message the agent speaks when a call connects | |
| number_of_words_for_interruption | No | Number of words the caller must speak to interrupt the agent |