> ## Documentation Index > Fetch the complete documentation index at: https://voice.ai/llms.txt > Use this file to discover all available pages before exploring further. # Generate Speech > Generate speech from text. If voice_id is provided, uses that voice; otherwise uses the default built-in voice. Returns complete audio file. Synchronous endpoint - blocks until generation completes. ## OpenAPI ````yaml /docs/openapi.json post /api/v1/tts/speech openapi: 3.1.0 info: title: Voice AI Developer API description: >- Generate realistic speech and clone voices. AI voice agents with retrieval augmented generation (RAG) and model context protocol (MCP) included. version: 1.5.0 contact: {} servers: - url: https://dev.voice.ai description: HTTP API endpoints security: - BearerAuth: [] tags: - name: Text To Speech description: Text-to-speech endpoints for voice cloning and speech generation. - name: Pronunciation Dictionaries description: Manage versioned pronunciation dictionaries for TTS and voice agents. - name: Agent Management description: Create, update, deploy, and manage voice agents. - name: Models description: Discover supported LLM and TTS model options. - name: Agent Connection description: Connect to agents and manage active calls. - name: Agent Analytics description: >- Analytics and reporting for agent performance, call history, and statistics. - name: Agent Knowledge Base description: >- Manage knowledge bases for retrieval augmented generation (RAG) with your agents. - name: Phone Number Management description: >- Search, select, and manage phone numbers from your plan allocation. Release numbers to free up slots. - name: Managed Tools description: >- Connect Google Calendar, Sheets, and Gmail to agents and inspect connection state. paths: /api/v1/tts/speech: post: tags: - Text To Speech summary: Generate Speech description: >- Generate speech from text. If voice_id is provided, uses that voice; otherwise uses the default built-in voice. Returns complete audio file. Synchronous endpoint - blocks until generation completes. operationId: generate_speech_api_v1_tts_speech_post requestBody: content: application/json: schema: $ref: '#/components/schemas/GenerateSpeechRequest' required: true responses: '200': description: Successful Response - Returns binary audio file (32kHz sample rate) content: audio/mpeg: schema: type: string format: binary description: MP3 audio file (32kHz sample rate, compressed) audio/wav: schema: type: string format: binary description: WAV audio file (32kHz sample rate, uncompressed with headers) audio/pcm: schema: type: string format: binary description: PCM audio (32kHz mono, 16-bit signed little-endian samples) '401': description: Unauthorized - Invalid or missing API key content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '402': description: Payment Required - Insufficient credits to generate speech content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '404': description: Not Found - Voice ID does not exist or is not accessible content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '422': description: Validation Error content: application/json: schema: $ref: '#/components/schemas/HTTPValidationError' '500': description: Internal Server Error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' components: schemas: GenerateSpeechRequest: properties: text: type: string title: Text description: >- The text to generate speech for. Supports inline spelling markup such as [SPELL:BX42] to spell acronyms, names, or codes character by character. voice_id: type: string title: Voice Id description: Voice ID to use. Omit to use the default built-in voice. audio_format: $ref: '#/components/schemas/AudioFormat' description: Audio output format (32kHz sample rate) default: mp3 temperature: type: number title: Temperature description: Sampling temperature (0.0-2.0) default: 1 minimum: 0 maximum: 2 top_p: type: number title: Top P description: Nucleus sampling parameter (0.0-1.0) default: 0.8 minimum: 0 maximum: 1 model: type: string title: Model description: >- TTS model to use. If not provided, automatically selected based on language. English uses non-multilingual models; other languages use multilingual models. enum: - voiceai-tts-v1-latest - voiceai-tts-v1-2026-02-10 - voiceai-tts-multilingual-v1-latest - voiceai-tts-multilingual-v1-2026-02-10 language: type: string title: Language description: Language code (ISO 639-1 format) default: en enum: - en - ca - sv - es - fr - de - it - pt - pl - ru - nl dictionary_id: type: string title: Dictionary Id description: Optional managed pronunciation dictionary identifier. dictionary_version: type: integer title: Dictionary Version description: >- Optional managed dictionary version. Defaults to latest when omitted. minimum: 1 type: object required: - text title: GenerateSpeechRequest ErrorResponse: properties: error: type: string title: Error description: Error message detail: type: string title: Detail description: Detailed error information nullable: true code: type: string title: Code description: Error code nullable: true type: object required: - error title: ErrorResponse HTTPValidationError: properties: detail: items: $ref: '#/components/schemas/ValidationError' type: array title: Detail type: object title: HTTPValidationError AudioFormat: type: string enum: - mp3 - wav - pcm - alaw_8000 - mp3_22050_32 - mp3_24000_48 - mp3_44100_32 - mp3_44100_64 - mp3_44100_96 - mp3_44100_128 - mp3_44100_192 - opus_48000_32 - opus_48000_64 - opus_48000_96 - opus_48000_128 - opus_48000_192 - pcm_8000 - pcm_16000 - pcm_22050 - pcm_24000 - pcm_32000 - pcm_44100 - pcm_48000 - ulaw_8000 - wav_16000 - wav_22050 - wav_24000 title: AudioFormat description: >- Audio output format. Basic formats (mp3, wav, pcm) use 32kHz sample rate. Format-specific options allow control over sample rate and bitrate. mp3/wav/pcm: 32kHz formats. mp3_*: MP3 with sample rate and bitrate (e.g., mp3_44100_128 = 44.1kHz, 128kbps). opus_*: Opus with sample rate and bitrate (e.g., opus_48000_64 = 48kHz, 64kbps). pcm_*: PCM with sample rate (e.g., pcm_16000 = 16kHz); all use 16-bit signed little-endian mono. wav_*: WAV with sample rate (e.g., wav_24000 = 24kHz). alaw_8000/ulaw_8000: Telephony formats (8kHz, A-law/μ-law). ValidationError: properties: loc: items: anyOf: - type: string - type: integer type: array title: Location msg: type: string title: Message type: type: string title: Error Type type: object required: - loc - msg - type title: ValidationError securitySchemes: BearerAuth: type: http scheme: bearer bearerFormat: JWT description: >- Bearer token authentication. Use your API key as the bearer token. Format: Authorization: Bearer ````