> ## Documentation Index > Fetch the complete documentation index at: https://docs.tryhamsa.com/llms.txt > Use this file to discover all available pages before exploring further. # Create Voice Agent Route > Create a new voice agent to start using our powerful and robust voice assistant feature. This route requires an API key for authentication. ## OpenAPI ````yaml /api-reference/openapi.json post /v1/voice-agents openapi: 3.0.1 info: title: Hamsa API description: Trying out our transcription and media-to-text APIs license: name: MIT version: 1.0.0 servers: - url: https://api.tryhamsa.com security: [] paths: /v1/voice-agents: post: summary: Create a new voice agent description: >- Create a new voice agent to start using our powerful and robust voice assistant feature. This route requires an API key for authentication. requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/VoiceAgentRequest' responses: '200': description: Successful response content: application/json: schema: $ref: '#/components/schemas/VoiceAgentResponse' '400': description: Bad Request content: application/json: schema: $ref: '#/components/schemas/ErrorSchema' '401': description: Bad Request content: application/json: schema: $ref: '#/components/schemas/ErrorSchema' '403': description: Bad Request content: application/json: schema: $ref: '#/components/schemas/ErrorSchema' '500': description: Internal Server Error content: application/json: schema: $ref: '#/components/schemas/ErrorSchema' deprecated: true security: - ApiKeyAuth: [] components: schemas: VoiceAgentRequest: type: object properties: agentName: type: string greetingMessage: description: >- Greeting Message can be customized by adding params like: {{name}}, to make the message more personal! type: string default: Hello, how can I help you today? greetingMessageType: type: string nullable: true enum: - STATIC - PROMPT default: null preamble: description: >- Preamble can be customized by adding params like: {{agentName}}, to make the prompt more specific and context aware! type: string default: You are a helpful assistant that will answer users questions. lang: type: string enum: - ar - en default: ar pokeMessages: description: The messages to say to the user when he doesn't respond! type: array items: type: string default: [] realTime: type: boolean default: false silenceThreshold: type: integer default: 800 minimum: 1 description: >- The silence threshold in milliseconds. Minimum value is 1 millisecond. interrupt: type: boolean default: true type: type: string default: Single Prompt outcome: description: >- A prompt to take the conversation and transform it to anything you like, like creating an article out of the conversation! type: string nullable: true default: null webhookUrl: type: string nullable: true format: uri example: https://example.com/webhook webhookAuth: $ref: '#/components/schemas/ExternalWebhookAuthScheme' outcomeResponseShape: type: object title: Outcome Response Schema description: >- Recursive OpenAPI-style shape with 'type', 'properties', 'required', 'items', ...etc. Please refer to the `OpenAPI Outcome Response Instructions` above for more information. example: type: object properties: test: type: number description: Hello world from Hamsa AI! required: - test voiceId: description: >- For trying voices and choosing one by yourself, or even cloning a voice, please visit Hamsa Agents platform voices tab [here](https://agents.tryhamsa.com/app/voices). type: string nullable: true default: null tools: type: object properties: genderDetection: type: boolean default: false smartCallEnd: type: boolean default: false default: genderDetection: false smartCallEnd: false params: description: >- Set the params you define in the greeting message or preamble fields here, the key is the same as the param and the value is the param value. Press 'Add new property' below to set the param name then press enter, after that add the value. type: object default: {} alignment: type: object properties: greetingMessage: type: string default: ltr preamble: type: string default: ltr default: greetingMessage: ltr preamble: ltr webToolsIds: type: array items: type: string voiceDictionaryIds: type: array items: type: string knowledgeBaseItemsIds: type: array items: type: string webToolsOverrides: type: object userInactivityTimeout: type: number minimum: 5 maximum: 30 default: 15 maxCallDuration: type: number minimum: 30 maximum: 600 default: 300 responseDelay: type: number minimum: 100 maximum: 1500 default: 400 backgroundNoise: type: boolean default: false waitForUserToSpeakFirst: type: integer minimum: 0 maximum: 2 default: 0 thinkingVoice: type: boolean default: false speakerIdentification: type: boolean default: false llmConfig: $ref: '#/components/schemas/VoiceAgentLlmConfig' noiseCancellation: type: string default: telephony-optimized cancelNoisePer: type: string default: conversation agenticRag: type: boolean default: false languageDialectSwitcher: type: boolean default: false minInterruptionDuration: type: number minimum: 0.2 maximum: 1.5 default: 0.5 vadActivationThreshold: type: number minimum: 0.2 maximum: 0.9 default: 0.5 enableAutoGainControl: type: boolean default: false sendDenoisedToStt: type: boolean default: false required: - agentName VoiceAgentResponse: type: object properties: success: type: boolean default: false message: type: string default: success data: $ref: '#/components/schemas/VoiceAgentResponseData' ErrorSchema: type: object properties: code: type: integer message: type: string ExternalWebhookAuthScheme: type: object description: >- Authorization header to be sent in this format: `Authorization: [authKey] [authSecret]` properties: authKey: type: string example: Token nullable: true authSecret: type: string example: Secret nullable: true VoiceAgentLlmConfig: type: object description: Configuration for the LLM (Large Language Model) used by the voice agent properties: provider: type: string enum: - OpenAI - Gemini - deepmyst - Custom - Groq description: The LLM provider to use default: OpenAI example: OpenAI modelName: type: string description: >- The model name to use. Valid models depend on the provider: OpenAI (GPT-5, GPT-5-Mini, GPT-5-Nano, GPT-4.1, GPT-4.1-Mini, GPT-4.1-Nano, GPT-4o, GPT-4o-mini), Gemini (Gemini 2.5-Pro, Gemini 2.5-Flash, gemini-3-flash-preview), deepmyst (gpt-4.1-optimize, gpt-4.1-mini-optimize), Groq (openai/gpt-oss-120b, openai/gpt-oss-20b) default: GPT-4.1 example: GPT-4.1 baseUrl: type: string description: Custom base URL for the LLM API. Required when provider is 'Custom' example: https://api.custom-llm.com/v1 apiKey: type: string description: >- API key for authentication with the LLM provider. Required when provider is 'Custom' example: sk-... temperature: type: number minimum: 0 maximum: 1 description: >- Controls randomness in the model's responses. Lower values make output more deterministic, higher values make it more creative. Note: GPT-5, GPT-5-Mini, and GPT-5-Nano models require temperature to be exactly 1. default: 0.2 example: 0.2 VoiceAgentResponseData: type: object properties: id: type: string agentName: type: string greetingMessage: type: string greetingMessageType: type: string enum: - STATIC - PROMPT nullable: true description: type: string preamble: type: string lang: type: string pokeMessages: type: array items: type: string realTime: type: boolean silenceThreshold: type: integer interrupt: type: boolean type: type: string outcome: type: string nullable: true outcomeResponseShape: type: object nullable: true example: type: object properties: test: type: number description: Hello world from Hamsa AI! required: - test projectId: type: string apiKeyId: type: string voiceId: type: string nullable: true voiceRecordId: type: string nullable: true voiceRecord: type: object description: The voice record object which is connected to this agent. nullable: true collectionId: type: string nullable: true isTemplate: type: boolean icon: type: string nullable: true webhookUrl: type: string nullable: true format: uri example: https://example.com/webhook webhookAuth: $ref: '#/components/schemas/ExternalWebhookAuthScheme' params: type: object description: The parameters that are created for this agent. alignment: type: object properties: greetingMessage: type: string preamble: type: string tools: type: object properties: genderDetection: type: boolean smartCallEnd: type: boolean webToolsIds: type: array items: type: string voiceDictionaryIds: type: array items: type: string knowledgeBaseItemsIds: type: array items: type: string webToolsOverrides: type: object userInactivityTimeout: type: number maxCallDuration: type: number responseDelay: type: number backgroundNoise: type: boolean waitForUserToSpeakFirst: type: integer thinkingVoice: type: boolean speakerIdentification: type: boolean llmConfig: $ref: '#/components/schemas/VoiceAgentLlmConfig' noiseCancellation: type: string cancelNoisePer: type: string nullable: true agenticRag: type: boolean languageDialectSwitcher: type: boolean minInterruptionDuration: type: number vadActivationThreshold: type: number enableAutoGainControl: type: boolean sendDenoisedToStt: type: boolean createdAt: type: string format: date-time updatedAt: type: string format: date-time securitySchemes: ApiKeyAuth: type: apiKey in: header name: Authorization description: >- Pass the API key in the Authorization header, You need to put Token keyword before the API key. e.g. 'Authorization: Token ' ````