> ## Documentation Index > Fetch the complete documentation index at: https://docs.closerx.ai/llms.txt > Use this file to discover all available pages before exploring further. # Update Agent > Updates the configuration of an existing AI call agent or LLM by its ID ## OpenAPI ````yaml PATCH /agent-llm/ openapi: 3.1.0 info: title: CloserX.ai API description: >- The CloserX.ai API provides programmatic access to manage AI-powered calling operations, including agent creation, campaign management, contact handling, telephony integration, and more. This API enables seamless automation and integration with CRM systems, telephony providers, and other third-party services for enhanced client communication and business operations. license: name: MIT version: 1.0.0 servers: - url: https://app.closerx.ai/api security: - api_key: [] paths: /agent-llm/: patch: description: Updates the configuration of an existing AI call agent or LLM by its ID parameters: - name: Company-Name in: header description: The name of the schema to be used for the request required: true schema: type: string - name: request_for in: query description: Specifies whether the request is for an agent or LLM required: true schema: type: string enum: - agent - llm example: agent - name: pk in: query description: Unique identifier for the AI call agent or LLM to be updated required: true schema: type: number example: 12 requestBody: description: Configuration details to update for the AI call agent or LLM content: application/json: schema: $ref: '#/components/schemas/UpdateAgentAndLLMRequest' required: true responses: '200': description: AI call agent or LLM update response content: application/json: schema: $ref: '#/components/schemas/UpdateAgentAndLLMResponse' '400': description: Unexpected error content: application/json: schema: $ref: '#/components/schemas/Error' components: schemas: UpdateAgentAndLLMRequest: type: object oneOf: - properties: agent: type: object properties: agent_name: type: string description: The name of the agent, used for reference. nullable: true example: Jarvis voice_id: type: string description: Unique voice ID used for the agent. example: 11labs-Adrian voice_temperature: type: number minimum: 0 maximum: 2 description: >- Controls voice stability for ElevenLabs voices. Lower is more stable, higher is more variant. Defaults to 1. example: 1 voice_speed: type: number minimum: 0.5 maximum: 2 description: >- Controls voice speed. Lower is slower, higher is faster. Defaults to 1. example: 1 volume: type: number minimum: 0 maximum: 2 description: >- Controls agent volume. Lower is quieter, higher is louder. Defaults to 1. example: 1 responsiveness: type: number minimum: 0 maximum: 1 description: >- Controls agent responsiveness. Lower is slower, higher is faster. Defaults to 1. example: 1 interruption_sensitivity: type: number minimum: 0 maximum: 1 description: >- Controls sensitivity to user interruptions. Lower is harder to interrupt, higher is easier. Defaults to 1. example: 1 pronunciation_dictionary: type: array description: >- A list of words/phrases and their pronunciation to guide audio synthesis. Currently only supported for English & 11labs voices. Set to null to remove. items: type: object properties: word: type: string description: >- The string of word/phrase to be annotated with pronunciation. example: actually alphabet: type: string description: The phonetic alphabet to be used for pronunciation. enum: - cmu - ipa phoneme: type: string description: >- Pronunciation of the word in the format of IPA/CMU pronunciation. example: ˈæktʃuəli nullable: true enable_backchannel: type: boolean description: >- Enables backchannel responses (e.g., 'yeah', 'uh-huh'). Defaults to false. example: true backchannel_frequency: type: number minimum: 0 maximum: 1 description: >- Controls backchannel frequency when enabled. Lower is less frequent, higher is more frequent. Defaults to 0.8. example: 0.9 backchannel_words: type: array items: type: string description: List of backchannel words. Defaults apply if not set. nullable: true example: - yeah - uh-huh vocab_specialization: type: string enum: - general - medical description: >- Vocabulary set for transcription (English only). Defaults to general. example: general allow_user_dtmf: type: boolean description: >- If set to true, DTMF input will be accepted and processed. If false, any DTMF input will be ignored. Default to true. example: true user_dtmf_options: type: object description: DTMF options for the agent. properties: digit_limit: type: number description: >- The maximum number of digits allowed in the user's DTMF input per turn. timeout_ms: type: number description: >- The time (in milliseconds) to wait for user DTMF input before timing out. termination_key: type: string description: >- A single key that signals the end of DTMF input (0–9, #, *). nullable: true denoising_mode: type: string enum: - noise-cancellation - noise-and-background-speech-cancellation description: Denoising mode for audio. Defaults to noise-cancellation. example: noise-cancellation reminder_trigger_ms: type: number minimum: 0 description: >- Triggers reminder after user silence duration (ms). Defaults to 10000. example: 10000 reminder_max_count: type: integer minimum: 0 description: >- Maximum number of reminders for unresponsive user. Defaults to 1. example: 2 ambient_sound: type: string enum: - coffee-shop - convention-hall - summer-outdoor - mountain-outdoor - static-noise - call-center description: >- Ambient sound for realistic call experience. Set to null to remove. nullable: true ambient_sound_volume: type: number minimum: 0 maximum: 2 description: >- Controls ambient sound volume. Lower is quieter, higher is louder. Defaults to 1. example: 1 language: type: string enum: - en-US - en-IN - en-GB - en-AU - en-NZ - de-DE - es-ES - es-419 - hi-IN - fr-FR - fr-CA - ja-JP - pt-PT - pt-BR - zh-CN - ru-RU - it-IT - ko-KR - nl-NL - nl-BE - pl-PL - tr-TR - th-TH - vi-VN - ro-RO - bg-BG - ca-ES - da-DK - fi-FI - el-GR - hu-HU - id-ID - no-NO - sk-SK - sv-SE - multi description: >- Language and dialect for speech recognition. Defaults to en-US. example: en-US boosted_keywords: type: array items: type: string description: Keywords to bias transcriber model. Set to null to remove. nullable: true example: - aicall - kroger normalize_for_speech: type: boolean description: Normalizes text to spoken form for consistent synthesis. example: true end_call_after_silence_ms: type: integer minimum: 10000 description: >- Ends call after user silence duration (ms). Defaults to 600000. example: 600000 max_call_duration_ms: type: integer minimum: 60000 maximum: 7200000 description: Maximum call duration (ms). Defaults to 3600000. example: 3600000 voicemail_option: type: object description: Voicemail detection settings. Set to null to disable. nullable: true properties: action: type: object properties: type: type: string description: >- Type of voicemail action (e.g., static_text, hangup). example: static_text text: type: string description: Text for voicemail prompt, if applicable. example: Please give us a callback tomorrow at 10am. analysis_data: type: array items: oneOf: - type: object properties: type: type: string const: string description: >- The data type of the analysis field, set to string. name: type: string description: The name of the analysis field. examples: - sentiment - key_topic - action_item description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Key topic discussed - Action item assigned examples: type: array items: type: string description: Example values for the analysis field. nullable: true examples: - - happy customer - urgent issue - - pricing - product feedback - - follow-up call - send email required: - type - name - description - type: object properties: type: type: string const: enum description: The data type of the analysis field, set to enum. name: type: string description: The name of the analysis field. examples: - sentiment - status - priority description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Status of the call - Priority level choices: type: array items: type: string description: Possible values for the analysis field. nullable: true examples: - - positive - negative - neutral - - successful - unsuccessful - pending - - high - medium - low required: - type - name - description - type: object properties: type: type: string const: boolean description: >- The data type of the analysis field, set to boolean. name: type: string description: The name of the analysis field. examples: - is_urgent - is_resolved description: type: string description: >- A description of what the analysis field represents. examples: - Whether the issue is urgent - Whether the issue is resolved required: - type - name - description - type: object properties: type: type: string const: number description: >- The data type of the analysis field, set to number. name: type: string description: The name of the analysis field. examples: - call_duration - customer_rating description: type: string description: >- A description of what the analysis field represents. examples: - Duration of the call in minutes - Customer satisfaction rating required: - type - name - description description: Configuration for a single post-call analysis field. description: >- Additional data to extract post-call. Available after call ends. nullable: true ring_duration_ms: type: integer minimum: 5000 maximum: 90000 description: Ring duration for calls (ms). Defaults to 30000. example: 30000 required: - agent - properties: llm: type: object properties: model: type: string enum: - gpt-4o - gpt-4o-mini - gpt-4.1 - gpt-4.1-mini - gpt-4.1-nano - claude-3.7-sonnet - claude-3.5-haiku - gemini-2.0-flash - gemini-2.0-flash-lite nullable: true description: >- Select the underlying text LLM. If not set, defaults to gpt-4o. s2s_model: type: string enum: - gpt-4o-realtime - gpt-4o-mini-realtime nullable: true description: >- Select the underlying speech-to-speech model. Can only set this or model, not both. model_temperature: type: number format: float minimum: 0 maximum: 1 description: >- Controls randomness of the response. Lower values are more deterministic, higher values are more random. Defaults to 0 if unset. model_high_priority: type: boolean description: >- Enables fast tier with high priority pool for lower latency. Defaults to false. general_prompt: type: string nullable: true description: >- General prompt appended to system prompt regardless of agent state. general_tools: type: array items: oneOf: - type: object properties: type: type: string const: end_call description: The tool type to end the call. name: type: string description: The name of the tool. example: end_call description: type: string description: A description of what the tool does. example: Terminates the ongoing call. required: - type - name - description - type: object properties: type: type: string const: call_transfer description: The tool type to transfer the call. name: type: string description: The name of the tool, must not contain spaces. example: transfer_call1 pattern: ^[^\s]+$ description: type: string description: A description of what the tool does. example: Transfers the call to another agent or number. transfer_destination: type: object oneOf: - type: object properties: type: type: string const: predefined description: >- Indicates a predefined phone number for transfer. number: type: string description: >- The phone number to transfer the call to in E.164 format. example: '+911111111111' pattern: ^\+[1-9]\d{1,14}$ required: - type - number description: >- Details for transferring to a predefined number. - type: object properties: type: type: string const: inferred description: >- Indicates dynamic routing for call transfer. prompt: type: string description: >- The prompt to determine the transfer destination. example: >- If the user wants to reach support, transfer to +1 (925) 222-4452; if the user wants to reach sales, transfer to +1 (925) 1111-3444 required: - type - prompt description: >- Details for dynamically determining the transfer destination. description: The destination details for the call transfer. nullable: true transfer_option: type: object oneOf: - type: object properties: type: type: string const: cold_transfer description: Indicates a cold transfer. show_transferee_as_caller: type: boolean description: >- Whether to display the transferee's number as the caller ID. example: false agent_detection_timeout_ms: type: number description: >- Time in milliseconds to wait before considering the transfer failed. example: 5000 minimum: 10000 maximum: 300000 on_hold_music: type: string description: >- Music to play while the caller is being transferred. enum: - none - ringtone - relaxing_sound - uplifting_beats example: none opt_out_human_detection: type: boolean description: >- Whether to skip human detection for the transfer. example: false opt_out_initial_message: type: boolean description: >- Whether to skip the AI saying 'Hello' after connecting the call. example: false required: - type - show_transferee_as_caller - agent_detection_timeout_ms - on_hold_music - opt_out_human_detection - opt_out_initial_message description: Options for cold transfer. - type: object properties: type: type: string const: warm_transfer description: Indicates a warm transfer. public_handoff_option: type: object oneOf: - type: object properties: type: type: string const: prompt description: Indicates a prompt-based handoff. prompt: type: string description: >- The prompt for the warm transfer public handoff. example: >- Please hold while I connect you to the support team. required: - type - prompt description: >- Prompt-based public handoff for warm transfer. - type: object properties: type: type: string const: static_message description: Indicates a static message handoff. message: type: string description: >- The static message for the warm transfer public handoff. example: Transferring your call now. required: - type - message description: >- Static message public handoff for warm transfer. description: Public handoff options for warm transfer. private_handoff_option: type: object oneOf: - type: object properties: type: type: string const: prompt description: >- Indicates a prompt-based private handoff. prompt: type: string description: >- The prompt for the warm transfer private handoff. example: >- This is a call regarding a support issue. required: - type - prompt description: >- Prompt-based private handoff for warm transfer. - type: object properties: type: type: string const: static_message description: >- Indicates a static message private handoff. message: type: string description: >- The static message for the warm transfer private handoff. example: Support call incoming. required: - type - message description: >- Static message private handoff for warm transfer. description: Private handoff options for warm transfer. agent_detection_timeout_ms: type: number description: >- Time in milliseconds to wait before considering the transfer failed. example: 5000 minimum: 10000 maximum: 300000 on_hold_music: type: string description: >- Music to play while the caller is being transferred. enum: - none - ringtone - relaxing_sound - uplifting_beats example: none opt_out_human_detection: type: boolean description: >- Whether to skip human detection for the transfer. example: false opt_out_initial_message: type: boolean description: >- Whether to skip the AI saying 'Hello' after connecting the call. example: false required: - type - public_handoff_option - private_handoff_option - agent_detection_timeout_ms - on_hold_music - opt_out_human_detection - opt_out_initial_message description: Options for warm transfer. description: Options for the call transfer type. nullable: true required: - type - name - description - transfer_destination - transfer_option - type: object properties: type: type: string const: book_appointment_cal description: >- The tool type to book an appointment in the calendar. name: type: string description: The name of the tool. example: Book_Calendar description: type: string description: A description of what the tool does. example: Book Calendar Availability cal_api_key: type: string description: The Cal.com API key for accessing the calendar. example: cal_api_1234567890 event_type_id: type: integer description: The ID of the event type to book in the calendar. example: 12345 timezone: type: string description: The timezone for the appointment. example: America/Los_Angeles required: - type - name - description - cal_api_key - event_type_id - timezone - type: object properties: type: type: string const: check_availability_cal description: The tool type to check calendar availability. name: type: string description: The name of the tool. example: check_availability description: type: string description: A description of what the tool does. example: Checks available time slots in the calendar. parameters: type: object properties: cal_api_key: type: string description: >- The Cal.com API key for accessing the calendar. example: cal_api_1234567890 event_type_id: type: integer description: >- The ID of the event type to check for availability. example: 12345 date: type: string format: date description: The date to check for availability. example: '2025-08-15' timezone: type: string description: The timezone for the availability check. example: America/Los_Angeles required: - cal_api_key - event_type_id - date - timezone required: - type - name - description - parameters - type: object properties: type: type: string const: press_digit description: The tool type to press a digit during a call. name: type: string description: The name of the tool. example: press_digit description: type: string description: A description of what the tool does. example: Simulates pressing a digit on the phone keypad. parameters: type: object properties: digit: type: string description: 'The digit to press (0-9, *, #).' example: '1' required: - digit nullable: true required: - type - name - description - type: object properties: type: type: string const: agent_swap description: The tool type to swap to another agent. name: type: string description: The name of the tool. example: agent_transfer description: type: string description: A description of what the tool does. example: Swaps the current agent with another agent. agent_id: type: string description: The ID of the agent to swap to. example: agent_456 execution_message_description: type: string description: >- The message to be spoken during the agent swap execution, if enabled. example: Please hold while I transfer you to another agent. nullable: true post_call_analysis_setting: type: string enum: - both_agents - only_destination_agent description: >- Determines which agents are included in post-call analysis. example: both_agents speak_during_execution: type: boolean description: >- Whether to speak the execution message during the agent swap. example: true required: - type - name - description - agent_id - post_call_analysis_setting - speak_during_execution - type: object properties: type: type: string const: extract_dynamic_variable description: The tool type to extract dynamic variables. name: type: string description: >- Name of the tool. Must be unique, consist of a-z, A-Z, 0-9, underscores, or dashes, max length 64, no spaces. description: type: string description: >- Describes what the tool does, including when to call it. variables: type: array items: oneOf: - type: object properties: type: type: string const: string description: >- The data type of the analysis field, set to string. name: type: string description: The name of the analysis field. examples: - sentiment - key_topic - action_item description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Key topic discussed - Action item assigned examples: type: array items: type: string description: Example values for the analysis field. nullable: true examples: - - happy customer - urgent issue - - pricing - product feedback - - follow-up call - send email required: - type - name - description - type: object properties: type: type: string const: enum description: >- The data type of the analysis field, set to enum. name: type: string description: The name of the analysis field. examples: - sentiment - status - priority description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Status of the call - Priority level choices: type: array items: type: string description: Possible values for the analysis field. nullable: true examples: - - positive - negative - neutral - - successful - unsuccessful - pending - - high - medium - low required: - type - name - description - type: object properties: type: type: string const: boolean description: >- The data type of the analysis field, set to boolean. name: type: string description: The name of the analysis field. examples: - is_urgent - is_resolved description: type: string description: >- A description of what the analysis field represents. examples: - Whether the issue is urgent - Whether the issue is resolved required: - type - name - description - type: object properties: type: type: string const: number description: >- The data type of the analysis field, set to number. name: type: string description: The name of the analysis field. examples: - call_duration - customer_rating description: type: string description: >- A description of what the analysis field represents. examples: - Duration of the call in minutes - Customer satisfaction rating required: - type - name - description required: - type - name - description - variables - type: object properties: type: type: string const: mcp description: The tool type for MCP. name: type: string description: Name of the MCP tool. description: type: string description: Description of the MCP tool. mcp_id: type: string description: Unique ID of the MCP. input_schema: type: object description: The input schema of the MCP tool. additionalProperties: type: string child_attributes_hidden: true required: - type - name - description - mcp_id - input_schema description: Configuration for a single tool available to the model. description: >- List of tools the model may call (e.g., end call, transfer call, or custom tools). nullable: true begin_message: type: string nullable: true description: >- First utterance said by the agent. If empty, agent waits for user to speak. If not set, LLM generates dynamically. required: - llm UpdateAgentAndLLMResponse: type: object oneOf: - properties: agent: type: object properties: agent_name: type: string description: The name of the agent, used for reference. nullable: true example: Jarvis voice_id: type: string description: Unique voice ID used for the agent. example: 11labs-Adrian voice_temperature: type: number minimum: 0 maximum: 2 description: >- Controls voice stability for ElevenLabs voices. Lower is more stable, higher is more variant. Defaults to 1. example: 1 voice_speed: type: number minimum: 0.5 maximum: 2 description: >- Controls voice speed. Lower is slower, higher is faster. Defaults to 1. example: 1 volume: type: number minimum: 0 maximum: 2 description: >- Controls agent volume. Lower is quieter, higher is louder. Defaults to 1. example: 1 responsiveness: type: number minimum: 0 maximum: 1 description: >- Controls agent responsiveness. Lower is slower, higher is faster. Defaults to 1. example: 1 interruption_sensitivity: type: number minimum: 0 maximum: 1 description: >- Controls sensitivity to user interruptions. Lower is harder to interrupt, higher is easier. Defaults to 1. example: 1 pronunciation_dictionary: type: array description: >- A list of words/phrases and their pronunciation to guide audio synthesis. Currently only supported for English & 11labs voices. Set to null to remove. items: type: object properties: word: type: string description: >- The string of word/phrase to be annotated with pronunciation. example: actually alphabet: type: string description: The phonetic alphabet to be used for pronunciation. enum: - cmu - ipa phoneme: type: string description: >- Pronunciation of the word in the format of IPA/CMU pronunciation. example: ˈæktʃuəli nullable: true enable_backchannel: type: boolean description: >- Enables backchannel responses (e.g., 'yeah', 'uh-huh'). Defaults to false. example: true backchannel_frequency: type: number minimum: 0 maximum: 1 description: >- Controls backchannel frequency when enabled. Lower is less frequent, higher is more frequent. Defaults to 0.8. example: 0.9 backchannel_words: type: array items: type: string description: List of backchannel words. Defaults apply if not set. nullable: true example: - yeah - uh-huh vocab_specialization: type: string enum: - general - medical description: >- Vocabulary set for transcription (English only). Defaults to general. example: general allow_user_dtmf: type: boolean description: >- If set to true, DTMF input will be accepted and processed. If false, any DTMF input will be ignored. Default to true. example: true user_dtmf_options: type: object description: DTMF options for the agent. properties: digit_limit: type: number description: >- The maximum number of digits allowed in the user's DTMF input per turn. timeout_ms: type: number description: >- The time (in milliseconds) to wait for user DTMF input before timing out. termination_key: type: string description: >- A single key that signals the end of DTMF input (0–9, #, *). nullable: true denoising_mode: type: string enum: - noise-cancellation - noise-and-background-speech-cancellation description: Denoising mode for audio. Defaults to noise-cancellation. example: noise-cancellation reminder_trigger_ms: type: number minimum: 0 description: >- Triggers reminder after user silence duration (ms). Defaults to 10000. example: 10000 reminder_max_count: type: integer minimum: 0 description: >- Maximum number of reminders for unresponsive user. Defaults to 1. example: 2 ambient_sound: type: string enum: - coffee-shop - convention-hall - summer-outdoor - mountain-outdoor - static-noise - call-center description: >- Ambient sound for realistic call experience. Set to null to remove. nullable: true ambient_sound_volume: type: number minimum: 0 maximum: 2 description: >- Controls ambient sound volume. Lower is quieter, higher is louder. Defaults to 1. example: 1 language: type: string enum: - en-US - en-IN - en-GB - en-AU - en-NZ - de-DE - es-ES - es-419 - hi-IN - fr-FR - fr-CA - ja-JP - pt-PT - pt-BR - zh-CN - ru-RU - it-IT - ko-KR - nl-NL - nl-BE - pl-PL - tr-TR - th-TH - vi-VN - ro-RO - bg-BG - ca-ES - da-DK - fi-FI - el-GR - hu-HU - id-ID - no-NO - sk-SK - sv-SE - multi description: >- Language and dialect for speech recognition. Defaults to en-US. example: en-US boosted_keywords: type: array items: type: string description: Keywords to bias transcriber model. Set to null to remove. nullable: true example: - aicall - kroger normalize_for_speech: type: boolean description: Normalizes text to spoken form for consistent synthesis. example: true end_call_after_silence_ms: type: integer minimum: 10000 description: >- Ends call after user silence duration (ms). Defaults to 600000. example: 600000 max_call_duration_ms: type: integer minimum: 60000 maximum: 7200000 description: Maximum call duration (ms). Defaults to 3600000. example: 3600000 voicemail_option: type: object description: Voicemail detection settings. Set to null to disable. nullable: true properties: action: type: object properties: type: type: string description: >- Type of voicemail action (e.g., static_text, hangup). example: static_text text: type: string description: Text for voicemail prompt, if applicable. example: Please give us a callback tomorrow at 10am. analysis_data: type: array items: oneOf: - type: object properties: type: type: string const: string description: >- The data type of the analysis field, set to string. name: type: string description: The name of the analysis field. examples: - sentiment - key_topic - action_item description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Key topic discussed - Action item assigned examples: type: array items: type: string description: Example values for the analysis field. nullable: true examples: - - happy customer - urgent issue - - pricing - product feedback - - follow-up call - send email required: - type - name - description - type: object properties: type: type: string const: enum description: The data type of the analysis field, set to enum. name: type: string description: The name of the analysis field. examples: - sentiment - status - priority description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Status of the call - Priority level choices: type: array items: type: string description: Possible values for the analysis field. nullable: true examples: - - positive - negative - neutral - - successful - unsuccessful - pending - - high - medium - low required: - type - name - description - type: object properties: type: type: string const: boolean description: >- The data type of the analysis field, set to boolean. name: type: string description: The name of the analysis field. examples: - is_urgent - is_resolved description: type: string description: >- A description of what the analysis field represents. examples: - Whether the issue is urgent - Whether the issue is resolved required: - type - name - description - type: object properties: type: type: string const: number description: >- The data type of the analysis field, set to number. name: type: string description: The name of the analysis field. examples: - call_duration - customer_rating description: type: string description: >- A description of what the analysis field represents. examples: - Duration of the call in minutes - Customer satisfaction rating required: - type - name - description description: Configuration for a single post-call analysis field. description: >- Additional data to extract post-call. Available after call ends. nullable: true ring_duration_ms: type: integer minimum: 5000 maximum: 90000 description: Ring duration for calls (ms). Defaults to 30000. example: 30000 required: - agent - properties: llm: type: object properties: model: type: string enum: - gpt-4o - gpt-4o-mini - gpt-4.1 - gpt-4.1-mini - gpt-4.1-nano - claude-3.7-sonnet - claude-3.5-haiku - gemini-2.0-flash - gemini-2.0-flash-lite nullable: true description: >- Select the underlying text LLM. If not set, defaults to gpt-4o. s2s_model: type: string enum: - gpt-4o-realtime - gpt-4o-mini-realtime nullable: true description: >- Select the underlying speech-to-speech model. Can only set this or model, not both. model_temperature: type: number format: float minimum: 0 maximum: 1 description: >- Controls randomness of the response. Lower values are more deterministic, higher values are more random. Defaults to 0 if unset. model_high_priority: type: boolean description: >- Enables fast tier with high priority pool for lower latency. Defaults to false. general_prompt: type: string nullable: true description: >- General prompt appended to system prompt regardless of agent state. general_tools: type: array items: oneOf: - type: object properties: type: type: string const: end_call description: The tool type to end the call. name: type: string description: The name of the tool. example: end_call description: type: string description: A description of what the tool does. example: Terminates the ongoing call. required: - type - name - description - type: object properties: type: type: string const: call_transfer description: The tool type to transfer the call. name: type: string description: The name of the tool, must not contain spaces. example: transfer_call1 pattern: ^[^\s]+$ description: type: string description: A description of what the tool does. example: Transfers the call to another agent or number. transfer_destination: type: object oneOf: - type: object properties: type: type: string const: predefined description: >- Indicates a predefined phone number for transfer. number: type: string description: >- The phone number to transfer the call to in E.164 format. example: '+911111111111' pattern: ^\+[1-9]\d{1,14}$ required: - type - number description: >- Details for transferring to a predefined number. - type: object properties: type: type: string const: inferred description: >- Indicates dynamic routing for call transfer. prompt: type: string description: >- The prompt to determine the transfer destination. example: >- If the user wants to reach support, transfer to +1 (925) 222-4452; if the user wants to reach sales, transfer to +1 (925) 1111-3444 required: - type - prompt description: >- Details for dynamically determining the transfer destination. description: The destination details for the call transfer. nullable: true transfer_option: type: object oneOf: - type: object properties: type: type: string const: cold_transfer description: Indicates a cold transfer. show_transferee_as_caller: type: boolean description: >- Whether to display the transferee's number as the caller ID. example: false agent_detection_timeout_ms: type: number description: >- Time in milliseconds to wait before considering the transfer failed. example: 5000 minimum: 10000 maximum: 300000 on_hold_music: type: string description: >- Music to play while the caller is being transferred. enum: - none - ringtone - relaxing_sound - uplifting_beats example: none opt_out_human_detection: type: boolean description: >- Whether to skip human detection for the transfer. example: false opt_out_initial_message: type: boolean description: >- Whether to skip the AI saying 'Hello' after connecting the call. example: false required: - type - show_transferee_as_caller - agent_detection_timeout_ms - on_hold_music - opt_out_human_detection - opt_out_initial_message description: Options for cold transfer. - type: object properties: type: type: string const: warm_transfer description: Indicates a warm transfer. public_handoff_option: type: object oneOf: - type: object properties: type: type: string const: prompt description: Indicates a prompt-based handoff. prompt: type: string description: >- The prompt for the warm transfer public handoff. example: >- Please hold while I connect you to the support team. required: - type - prompt description: >- Prompt-based public handoff for warm transfer. - type: object properties: type: type: string const: static_message description: Indicates a static message handoff. message: type: string description: >- The static message for the warm transfer public handoff. example: Transferring your call now. required: - type - message description: >- Static message public handoff for warm transfer. description: Public handoff options for warm transfer. private_handoff_option: type: object oneOf: - type: object properties: type: type: string const: prompt description: >- Indicates a prompt-based private handoff. prompt: type: string description: >- The prompt for the warm transfer private handoff. example: >- This is a call regarding a support issue. required: - type - prompt description: >- Prompt-based private handoff for warm transfer. - type: object properties: type: type: string const: static_message description: >- Indicates a static message private handoff. message: type: string description: >- The static message for the warm transfer private handoff. example: Support call incoming. required: - type - message description: >- Static message private handoff for warm transfer. description: Private handoff options for warm transfer. agent_detection_timeout_ms: type: number description: >- Time in milliseconds to wait before considering the transfer failed. example: 5000 minimum: 10000 maximum: 300000 on_hold_music: type: string description: >- Music to play while the caller is being transferred. enum: - none - ringtone - relaxing_sound - uplifting_beats example: none opt_out_human_detection: type: boolean description: >- Whether to skip human detection for the transfer. example: false opt_out_initial_message: type: boolean description: >- Whether to skip the AI saying 'Hello' after connecting the call. example: false required: - type - public_handoff_option - private_handoff_option - agent_detection_timeout_ms - on_hold_music - opt_out_human_detection - opt_out_initial_message description: Options for warm transfer. description: Options for the call transfer type. nullable: true required: - type - name - description - transfer_destination - transfer_option - type: object properties: type: type: string const: book_appointment_cal description: >- The tool type to book an appointment in the calendar. name: type: string description: The name of the tool. example: Book_Calendar description: type: string description: A description of what the tool does. example: Book Calendar Availability cal_api_key: type: string description: The Cal.com API key for accessing the calendar. example: cal_api_1234567890 event_type_id: type: integer description: The ID of the event type to book in the calendar. example: 12345 timezone: type: string description: The timezone for the appointment. example: America/Los_Angeles required: - type - name - description - cal_api_key - event_type_id - timezone - type: object properties: type: type: string const: check_availability_cal description: The tool type to check calendar availability. name: type: string description: The name of the tool. example: check_availability description: type: string description: A description of what the tool does. example: Checks available time slots in the calendar. parameters: type: object properties: cal_api_key: type: string description: >- The Cal.com API key for accessing the calendar. example: cal_api_1234567890 event_type_id: type: integer description: >- The ID of the event type to check for availability. example: 12345 date: type: string format: date description: The date to check for availability. example: '2025-08-15' timezone: type: string description: The timezone for the availability check. example: America/Los_Angeles required: - cal_api_key - event_type_id - date - timezone required: - type - name - description - parameters - type: object properties: type: type: string const: press_digit description: The tool type to press a digit during a call. name: type: string description: The name of the tool. example: press_digit description: type: string description: A description of what the tool does. example: Simulates pressing a digit on the phone keypad. parameters: type: object properties: digit: type: string description: 'The digit to press (0-9, *, #).' example: '1' required: - digit nullable: true required: - type - name - description - type: object properties: type: type: string const: agent_swap description: The tool type to swap to another agent. name: type: string description: The name of the tool. example: agent_transfer description: type: string description: A description of what the tool does. example: Swaps the current agent with another agent. agent_id: type: string description: The ID of the agent to swap to. example: agent_456 execution_message_description: type: string description: >- The message to be spoken during the agent swap execution, if enabled. example: Please hold while I transfer you to another agent. nullable: true post_call_analysis_setting: type: string enum: - both_agents - only_destination_agent description: >- Determines which agents are included in post-call analysis. example: both_agents speak_during_execution: type: boolean description: >- Whether to speak the execution message during the agent swap. example: true required: - type - name - description - agent_id - post_call_analysis_setting - speak_during_execution - type: object properties: type: type: string const: extract_dynamic_variable description: The tool type to extract dynamic variables. name: type: string description: >- Name of the tool. Must be unique, consist of a-z, A-Z, 0-9, underscores, or dashes, max length 64, no spaces. description: type: string description: >- Describes what the tool does, including when to call it. variables: type: array items: oneOf: - type: object properties: type: type: string const: string description: >- The data type of the analysis field, set to string. name: type: string description: The name of the analysis field. examples: - sentiment - key_topic - action_item description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Key topic discussed - Action item assigned examples: type: array items: type: string description: Example values for the analysis field. nullable: true examples: - - happy customer - urgent issue - - pricing - product feedback - - follow-up call - send email required: - type - name - description - type: object properties: type: type: string const: enum description: >- The data type of the analysis field, set to enum. name: type: string description: The name of the analysis field. examples: - sentiment - status - priority description: type: string description: >- A description of what the analysis field represents. examples: - Sentiment of the call - Status of the call - Priority level choices: type: array items: type: string description: Possible values for the analysis field. nullable: true examples: - - positive - negative - neutral - - successful - unsuccessful - pending - - high - medium - low required: - type - name - description - type: object properties: type: type: string const: boolean description: >- The data type of the analysis field, set to boolean. name: type: string description: The name of the analysis field. examples: - is_urgent - is_resolved description: type: string description: >- A description of what the analysis field represents. examples: - Whether the issue is urgent - Whether the issue is resolved required: - type - name - description - type: object properties: type: type: string const: number description: >- The data type of the analysis field, set to number. name: type: string description: The name of the analysis field. examples: - call_duration - customer_rating description: type: string description: >- A description of what the analysis field represents. examples: - Duration of the call in minutes - Customer satisfaction rating required: - type - name - description required: - type - name - description - variables - type: object properties: type: type: string const: mcp description: The tool type for MCP. name: type: string description: Name of the MCP tool. description: type: string description: Description of the MCP tool. mcp_id: type: string description: Unique ID of the MCP. input_schema: type: object description: The input schema of the MCP tool. additionalProperties: type: string child_attributes_hidden: true required: - type - name - description - mcp_id - input_schema description: Configuration for a single tool available to the model. description: >- List of tools the model may call (e.g., end call, transfer call, or custom tools). nullable: true begin_message: type: string nullable: true description: >- First utterance said by the agent. If empty, agent waits for user to speak. If not set, LLM generates dynamically. required: - llm Error: required: - success - message - data type: object properties: success: type: boolean message: type: string data: type: object securitySchemes: api_key: type: apiKey in: header name: api_key ````