> ## Documentation Index > Fetch the complete documentation index at: https://docs.vogent.ai/llms.txt > Use this file to discover all available pages before exploring further. # Websocket TTS > Real-time text-to-speech streaming over websockets. Send text incrementally and receive audio chunks as they are generated. To authenticate, you must include an `Authorization` header, with the value `Bearer {API_TOKEN}`. You can also include an `apiKey` query param with the value being your API key. Websockets can run multiple TTS streams concurrently. If you're using the `contextId` feature, make sure all generations for the context are on the same websocket. ## AsyncAPI ````yaml api-reference/asyncapi.yaml channelOne id: channelOne title: Vogent Websocket description: >- Real-time text-to-speech streaming over websockets. Send text incrementally and receive audio chunks as they are generated. To authenticate, you must include an `Authorization` header, with the value `Bearer {API_TOKEN}`. You can also include an `apiKey` query param with the value being your API key. Websockets can run multiple TTS streams concurrently. If you're using the `contextId` feature, make sure all generations for the context are on the same websocket. servers: - id: vogent protocol: wss host: api.vogent.ai/api/tts/websocket bindings: [] variables: [] address: / parameters: [] bindings: [] operations: - &ref_1 id: sendWebsocketInput title: Send websocket input type: receive messages: - &ref_3 id: WebsocketInput payload: - name: Websocket Input description: Input message for text-to-speech generation type: object properties: - name: generationId type: string description: Unique identifier for the generation request. required: true - name: contextId type: string description: > A context that can represent a single conversation, helps preserve voice consistency across multiple lines. Leave blank if you don't need this behavior. Ignored after the first time this generationId is seen. You may not have different voices across the same context (all generations with the same contextId, must also have the same voiceId). required: false - name: items type: string required: false - name: description type: string description: > **Ignore the type above, this is an array.** An array of {"optionId": string, "value": string} objects (a ModelOptionValue object) required: false - name: voiceId type: string description: >- ID of the voice to use for generation. Required the first time this generationId is sent. required: false - name: text type: string description: Text content to convert to speech. required: false - name: finalText type: boolean description: Whether this is the final text chunk. required: false - name: cancel type: boolean description: Whether to cancel the generation with generationId. required: false - name: sampleRate type: integer description: >- Sample rate for the generated audio. Required the first time this generationId is sent. required: false - name: addWordTimestamps type: boolean description: Add timestamps on the audio for each word in the input. required: false headers: [] jsonPayloadSchema: type: object properties: generationId: type: string description: Unique identifier for the generation request. x-parser-schema-id: contextId: type: string description: > A context that can represent a single conversation, helps preserve voice consistency across multiple lines. Leave blank if you don't need this behavior. Ignored after the first time this generationId is seen. You may not have different voices across the same context (all generations with the same contextId, must also have the same voiceId). x-parser-schema-id: voiceOptionValues: items: type: string description: > **Ignore the type above, this is an array.** An array of {"optionId": string, "value": string} objects (a ModelOptionValue object) x-parser-schema-id: voiceId: type: string description: >- ID of the voice to use for generation. Required the first time this generationId is sent. x-parser-schema-id: text: type: string description: Text content to convert to speech. x-parser-schema-id: finalText: type: boolean description: Whether this is the final text chunk. x-parser-schema-id: cancel: type: boolean description: Whether to cancel the generation with generationId. x-parser-schema-id: sampleRate: type: integer description: >- Sample rate for the generated audio. Required the first time this generationId is sent. x-parser-schema-id: addWordTimestamps: type: boolean description: Add timestamps on the audio for each word in the input. x-parser-schema-id: required: - generationId x-parser-schema-id: title: Websocket Input description: Input message for text-to-speech generation example: |- { "generationId": "", "contextId": "", "items": "", "description": "", "voiceId": "", "text": "", "finalText": true, "cancel": true, "sampleRate": 123, "addWordTimestamps": true } bindings: [] extensions: - id: x-parser-unique-object-id value: WebsocketInput - id: x-parser-message-name value: WebsocketInput bindings: [] extensions: &ref_0 - id: x-parser-unique-object-id value: channelOne - &ref_2 id: receiveWebsocketOutput title: Receive websocket output type: send messages: - &ref_4 id: FinishedMessage payload: - name: Finished Message description: Message sent when text-to-speech generation is complete type: object properties: - name: type type: string description: Message type, always "finished" enumValues: - finished required: true - name: generationId type: string description: Unique identifier for the generation request required: true headers: [] jsonPayloadSchema: type: object properties: type: type: string enum: - finished description: Message type, always "finished" x-parser-schema-id: generationId: type: string description: Unique identifier for the generation request x-parser-schema-id: required: - type - generationId x-parser-schema-id: title: Finished Message description: Message sent when text-to-speech generation is complete example: |- { "type": "", "generationId": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: FinishedMessage - id: x-parser-message-name value: FinishedMessage - &ref_5 id: ErrorMessage payload: - name: Error Message description: Message sent when text-to-speech generation encounters an error type: object properties: - name: type type: string description: Message type, always "error" enumValues: - error required: true - name: generationId type: string description: Unique identifier for the generation request required: true - name: error type: string description: Error message describing what went wrong required: true headers: [] jsonPayloadSchema: type: object properties: type: type: string enum: - error description: Message type, always "error" x-parser-schema-id: generationId: type: string description: Unique identifier for the generation request x-parser-schema-id: error: type: string description: Error message describing what went wrong x-parser-schema-id: required: - type - generationId - error x-parser-schema-id: title: Error Message description: Message sent when text-to-speech generation encounters an error example: |- { "type": "", "generationId": "", "error": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: ErrorMessage - id: x-parser-message-name value: ErrorMessage - &ref_6 id: ChunkMessage payload: - name: Chunk Message description: >- Message sent with audio data chunks during text-to-speech generation type: object properties: - name: type type: string description: Message type, always "chunk" enumValues: - chunk required: true - name: generationId type: string description: Unique identifier for the generation request required: true - name: audio type: string description: Base64-encoded audio data chunk required: true headers: [] jsonPayloadSchema: type: object properties: type: type: string enum: - chunk description: Message type, always "chunk" x-parser-schema-id: generationId: type: string description: Unique identifier for the generation request x-parser-schema-id: audio: type: string description: Base64-encoded audio data chunk x-parser-schema-id: required: - type - generationId - audio x-parser-schema-id: title: Chunk Message description: Message sent with audio data chunks during text-to-speech generation example: |- { "type": "", "generationId": "", "audio": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: ChunkMessage - id: x-parser-message-name value: ChunkMessage - &ref_7 id: TimestampMessage payload: - name: Timestamp Message description: >- Message sent with timestamps during text-to-speech generation, only when add word timestamps is true type: object properties: - name: alignment type: object required: true properties: - name: words type: array description: >- An array of strings representing the words in this alignment chunk required: false properties: - name: item type: string required: false - name: startTimes type: array description: >- The start time of each word (the array will be the same length as words). Relative to the start of the generation, in milliseconds. required: false properties: - name: item type: integer required: false - name: durationMs type: array description: >- The duration of each word (the array will be the same length as words) required: false properties: - name: item type: integer required: false - name: type type: string description: Message type, always "alignment" enumValues: - alignment required: true - name: generationId type: string description: Unique identifier for the generation request required: true headers: [] jsonPayloadSchema: type: object properties: alignment: type: object properties: words: type: array items: type: string x-parser-schema-id: description: >- An array of strings representing the words in this alignment chunk x-parser-schema-id: startTimes: type: array items: type: integer x-parser-schema-id: description: >- The start time of each word (the array will be the same length as words). Relative to the start of the generation, in milliseconds. x-parser-schema-id: durationMs: type: array items: type: integer x-parser-schema-id: description: >- The duration of each word (the array will be the same length as words) x-parser-schema-id: x-parser-schema-id: type: type: string enum: - alignment description: Message type, always "alignment" x-parser-schema-id: generationId: type: string description: Unique identifier for the generation request x-parser-schema-id: required: - type - alignment - generationId x-parser-schema-id: title: Timestamp Message description: >- Message sent with timestamps during text-to-speech generation, only when add word timestamps is true example: No examples found bindings: [] extensions: - id: x-parser-unique-object-id value: TimestampMessage - id: x-parser-message-name value: TimestampMessage bindings: [] extensions: *ref_0 sendOperations: - *ref_1 receiveOperations: - *ref_2 sendMessages: - *ref_3 receiveMessages: - *ref_4 - *ref_5 - *ref_6 - *ref_7 extensions: - id: x-parser-unique-object-id value: channelOne securitySchemes: [] ````