> ## Documentation Index > Fetch the complete documentation index at: https://docs.inworld.ai/llms.txt > Use this file to discover all available pages before exploring further. # Synthesize speech (WebSocket) > Generate audio from text input while managing multiple independent audio generation streams over a single WebSocket connection. The independent audio streams each correspond to a *context*, identified by `contextId`, that maintains its own state. To use the API: - Create a context with audio generation configurations. By default, we allow up to 20 concurrent connections, with a maximum of 5 contexts per connection. - When you send text to be synthesized into audio, you can send it to a specific context (optional if there is only 1 context). - Each context maintains its own buffer that can be flushed either manually or automatically when the buffer reaches a certain threshold (see `maxBufferDelayMs` and `bufferCharThreshold` in the context configurations). - If texts are sent in full sentences phrases, it's recommended to use `auto_mode` which would automatically balance latency and quality of the generations. - Responses contain the `contextId` so you can match the audio to the request. - Close a context when it is no longer needed. ## AsyncAPI ````yaml api-reference/ttsAPI/synthesize-speech-websocket.json ttsStream id: ttsStream title: Tts stream description: '' servers: - id: production protocol: wss host: api.inworld.ai bindings: [] variables: [] address: /tts/v1/voice:streamBidirectional parameters: [] bindings: [] operations: - &ref_13 id: sendRequest title: Send request type: receive messages: - &ref_15 id: createContext payload: - name: Create Context description: >- Create a new context with specified voice and configuration. A context is an independent conversation happening over the connection. The configurations for each context are completely separate – you can have different voice ids, models, output formats, etc. between contexts. *Note*: for each connection, 5 contexts is the max. If you don't need multiple contexts, you can omit the contextId in the message to use a single context connection. type: object properties: - name: create type: object required: true properties: - name: voiceId type: string description: The identifier of the voice to use for the synthesis required: false - name: modelId type: string description: >- The ID of the model to use for synthesizing speech. See [Models](../../../tts/tts-models) for available models. required: false - name: audioConfig type: object required: false properties: - name: audioEncoding type: string description: >- The desired output format of the synthesized audio. Defaults to `MP3`. - `LINEAR16`: Uncompressed 16-bit signed little-endian samples (Linear PCM). The WAV header is included in every audio chunk. - `MP3`: MP3 audio. - `OGG_OPUS`: Opus encoded audio wrapped in an ogg container. The result will be a file which can be played natively on Android, and in browsers (at least Chrome and Firefox). The quality of the encoding is considerably higher than MP3 while using approximately the same bitrate. - `ALAW`: ALAW encoded audio. 8-bit companded PCM. - `MULAW`: MULAW encoded audio. 8-bit companded PCM. - `PCM`: PCM audio. Uncompressed 16-bit signed little-endian samples with no WAV header. - `WAV`: WAV audio. Uncompressed 16-bit signed little-endian samples. The WAV header is included in the first audio chunk only. On each flush_completed response, the next audio chunk will also start with a header. enumValues: - AUDIO_ENCODING_UNSPECIFIED - LINEAR16 - MP3 - OGG_OPUS - ALAW - MULAW - PCM - WAV required: false - name: sampleRateHertz type: integer description: >- The synthesis sample rate (in hertz) for this audio. Accepts values within the range [8000, 48000]. When this is specified, if this is different from the voice's natural sample rate, then the audio will be converted to the desired sample rate (which might result in worse audio quality), unless the specified sample rate is not supported for the encoding chosen, in which case it will fail the request and return an error. The default is 48,000. required: false - name: bitRate type: integer description: >- Bits per second of the audio. Only for compressed audio formats (`MP3`, `OGG_OPUS`). The default is 128,000. required: false - name: speakingRate type: number description: >- Speaking rate/speed, in the range [0.5, 1.5]. The default is 1.0, which is the normal native speed supported by the specific voice. We recommend using values above 0.8 to ensure high quality. required: false - name: temperature type: number description: >- *Ignored on `inworld-tts-2`. Use [`deliveryMode`](#deliveryMode) instead.* Determines the degree of randomness when sampling audio tokens to generate the response. Defaults to 1.0. Accepts values between 0 (exclusive) and 2 (inclusive). Higher values will make the output more random and can lead to more expressive results. Lower values will make it more deterministic. If 0 is provided, the default value will be used. For the most stable results, we recommend using the default value. required: false - name: timestampType type: string description: >- Controls timestamp metadata returned with the audio. When enabled, the response includes timing arrays, which can be useful for word-highlighting, karaoke-style captions, and lipsync. - WORD: Output arrays under `timestampInfo.wordAlignment` (words, wordStartTimeSeconds, wordEndTimeSeconds). - CHARACTER: Output arrays under `timestampInfo.characterAlignment` (characters, characterStartTimeSeconds, characterEndTimeSeconds). - TIMESTAMP_TYPE_UNSPECIFIED: Do not compute alignment; timestamp arrays will be empty or omitted. **Latency note:** Alignment adds additional computation. Enabling alignment can increase latency. The timestamps reset per flush, either triggered manually or automatically by the server. When you receive a `flushCompleted` message, the timestamps for subsequent chunks will start from 0. enumValues: - TIMESTAMP_TYPE_UNSPECIFIED - WORD - CHARACTER required: false - name: maxBufferDelayMs type: integer description: >- If set, determines the maximum time in milliseconds to buffer before starting generation. The timer starts running when the first text in the buffer is received and resets when new text arrives. If not set or set to 0, there will be no time-based limit on the buffer. Instead, the user will need to flush the buffer or set the bufferCharThreshold to trigger audio generation. Note that both length and timeout based flushing can be used together - if both maxBufferDelayMs and bufferCharThreshold are set, the server will flush if either condition is met. required: false - name: bufferCharThreshold type: integer description: >- Defines the minimum number of characters in the buffer that would automatically trigger audio generation. This allows you to rely on automatic triggering instead of calling flush manually. If set to 0 or left undefined, the threshold defaults to 1000 to ensure stable behavior. Cannot be set to a value greater than 1000. Note that the buffer will automatically flush if more than 1000 characters are accumulated, regardless of this setting. required: false - name: applyTextNormalization type: string description: >- When enabled, text normalization automatically expands and standardizes things like numbers, dates, times, and abbreviations before converting them to speech. For example, Dr. Smith becomes Doctor Smith, and 3/10/25 is spoken as March tenth, twenty twenty-five. Turning this off may reduce latency, but the speech output will read the text exactly as written. Defaults to automatically deciding whether to apply text normalization. enumValues: - APPLY_TEXT_NORMALIZATION_UNSPECIFIED - 'ON' - 'OFF' required: false - name: autoMode type: boolean description: >- When enabled, the server will control flushing of buffered text to achieve minimal latency, while still maintaining high quality audio output. Recommended when texts are sent in full sentences/phrases. The default is false. required: false - name: timestampTransportStrategy type: string description: >- The transport strategy of timestamps info. - `TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED`: The service will automatically decide the transport strategy. - `SYNC`: Timestamps will be returned in the same message as the audio data. - `ASYNC`: Timestamps could return in trailing message after the audio data. Use this strategy to reduce latency of the first audio chunk with v1.5+ models. enumValues: - TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED - SYNC - ASYNC required: false - name: language type: string description: >- BCP-47 language tag (e.g., `en-US`, `fr-FR`, `ja-JP`) specifying the language that the given voice should speak the text in. If a localized voice prompt exists for the language, it will be used. When omitted, the original voice prompt will be used and the language will be auto-detected from the input text. If an invalid language code is provided, an error will be returned. See [Languages](../../../tts/capabilities/multilingual) for more details. required: false - name: deliveryMode type: string description: >- *Only supported by `inworld-tts-2`. The field is ignored on other models.* Controls how varied the output is. - `DELIVERY_MODE_UNSPECIFIED`: Defaults to `BALANCED` behavior. - `STABLE`: Optimizes for more consistent, predictable output. - `BALANCED`: Balanced between stability and diversity. - `CREATIVE`: Optimizes for increased emotional range and variation. enumValues: - DELIVERY_MODE_UNSPECIFIED - STABLE - BALANCED - CREATIVE required: false - name: contextId type: string description: Optional context ID. If not provided, one will be generated required: false headers: [] jsonPayloadSchema: type: object properties: create: type: object properties: voiceId: &ref_0 type: string description: The identifier of the voice to use for the synthesis x-parser-schema-id: VoiceId modelId: type: string description: >- The ID of the model to use for synthesizing speech. See [Models](../../../tts/tts-models) for available models. x-parser-schema-id: audioConfig: &ref_1 type: object properties: audioEncoding: type: string enum: - AUDIO_ENCODING_UNSPECIFIED - LINEAR16 - MP3 - OGG_OPUS - ALAW - MULAW - PCM - WAV default: MP3 description: >- The desired output format of the synthesized audio. Defaults to `MP3`. - `LINEAR16`: Uncompressed 16-bit signed little-endian samples (Linear PCM). The WAV header is included in every audio chunk. - `MP3`: MP3 audio. - `OGG_OPUS`: Opus encoded audio wrapped in an ogg container. The result will be a file which can be played natively on Android, and in browsers (at least Chrome and Firefox). The quality of the encoding is considerably higher than MP3 while using approximately the same bitrate. - `ALAW`: ALAW encoded audio. 8-bit companded PCM. - `MULAW`: MULAW encoded audio. 8-bit companded PCM. - `PCM`: PCM audio. Uncompressed 16-bit signed little-endian samples with no WAV header. - `WAV`: WAV audio. Uncompressed 16-bit signed little-endian samples. The WAV header is included in the first audio chunk only. On each flush_completed response, the next audio chunk will also start with a header. x-parser-schema-id: sampleRateHertz: type: integer format: int32 description: >- The synthesis sample rate (in hertz) for this audio. Accepts values within the range [8000, 48000]. When this is specified, if this is different from the voice's natural sample rate, then the audio will be converted to the desired sample rate (which might result in worse audio quality), unless the specified sample rate is not supported for the encoding chosen, in which case it will fail the request and return an error. The default is 48,000. x-parser-schema-id: bitRate: type: integer format: int32 description: >- Bits per second of the audio. Only for compressed audio formats (`MP3`, `OGG_OPUS`). The default is 128,000. x-parser-schema-id: speakingRate: type: number format: double description: >- Speaking rate/speed, in the range [0.5, 1.5]. The default is 1.0, which is the normal native speed supported by the specific voice. We recommend using values above 0.8 to ensure high quality. x-parser-schema-id: x-parser-schema-id: AudioConfig temperature: &ref_2 type: number format: double description: >- *Ignored on `inworld-tts-2`. Use [`deliveryMode`](#deliveryMode) instead.* Determines the degree of randomness when sampling audio tokens to generate the response. Defaults to 1.0. Accepts values between 0 (exclusive) and 2 (inclusive). Higher values will make the output more random and can lead to more expressive results. Lower values will make it more deterministic. If 0 is provided, the default value will be used. For the most stable results, we recommend using the default value. x-parser-schema-id: Temperature timestampType: &ref_3 type: string enum: - TIMESTAMP_TYPE_UNSPECIFIED - WORD - CHARACTER default: TIMESTAMP_TYPE_UNSPECIFIED description: >- Controls timestamp metadata returned with the audio. When enabled, the response includes timing arrays, which can be useful for word-highlighting, karaoke-style captions, and lipsync. - WORD: Output arrays under `timestampInfo.wordAlignment` (words, wordStartTimeSeconds, wordEndTimeSeconds). - CHARACTER: Output arrays under `timestampInfo.characterAlignment` (characters, characterStartTimeSeconds, characterEndTimeSeconds). - TIMESTAMP_TYPE_UNSPECIFIED: Do not compute alignment; timestamp arrays will be empty or omitted. **Latency note:** Alignment adds additional computation. Enabling alignment can increase latency. The timestamps reset per flush, either triggered manually or automatically by the server. When you receive a `flushCompleted` message, the timestamps for subsequent chunks will start from 0. x-parser-schema-id: TimestampType maxBufferDelayMs: &ref_4 type: integer format: int32 description: >- If set, determines the maximum time in milliseconds to buffer before starting generation. The timer starts running when the first text in the buffer is received and resets when new text arrives. If not set or set to 0, there will be no time-based limit on the buffer. Instead, the user will need to flush the buffer or set the bufferCharThreshold to trigger audio generation. Note that both length and timeout based flushing can be used together - if both maxBufferDelayMs and bufferCharThreshold are set, the server will flush if either condition is met. x-parser-schema-id: MaxBufferDelayMs bufferCharThreshold: &ref_5 type: integer format: int32 description: >- Defines the minimum number of characters in the buffer that would automatically trigger audio generation. This allows you to rely on automatic triggering instead of calling flush manually. If set to 0 or left undefined, the threshold defaults to 1000 to ensure stable behavior. Cannot be set to a value greater than 1000. Note that the buffer will automatically flush if more than 1000 characters are accumulated, regardless of this setting. x-parser-schema-id: BufferCharThreshold applyTextNormalization: &ref_6 type: string enum: - APPLY_TEXT_NORMALIZATION_UNSPECIFIED - 'ON' - 'OFF' default: APPLY_TEXT_NORMALIZATION_UNSPECIFIED description: >- When enabled, text normalization automatically expands and standardizes things like numbers, dates, times, and abbreviations before converting them to speech. For example, Dr. Smith becomes Doctor Smith, and 3/10/25 is spoken as March tenth, twenty twenty-five. Turning this off may reduce latency, but the speech output will read the text exactly as written. Defaults to automatically deciding whether to apply text normalization. x-parser-schema-id: ApplyTextNormalization autoMode: &ref_7 type: boolean default: false description: >- When enabled, the server will control flushing of buffered text to achieve minimal latency, while still maintaining high quality audio output. Recommended when texts are sent in full sentences/phrases. The default is false. x-parser-schema-id: AutoMode timestampTransportStrategy: &ref_8 type: string enum: - TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED - SYNC - ASYNC default: TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED description: >- The transport strategy of timestamps info. - `TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED`: The service will automatically decide the transport strategy. - `SYNC`: Timestamps will be returned in the same message as the audio data. - `ASYNC`: Timestamps could return in trailing message after the audio data. Use this strategy to reduce latency of the first audio chunk with v1.5+ models. x-parser-schema-id: TimestampTransportStrategy language: &ref_9 type: string description: >- BCP-47 language tag (e.g., `en-US`, `fr-FR`, `ja-JP`) specifying the language that the given voice should speak the text in. If a localized voice prompt exists for the language, it will be used. When omitted, the original voice prompt will be used and the language will be auto-detected from the input text. If an invalid language code is provided, an error will be returned. See [Languages](../../../tts/capabilities/multilingual) for more details. x-parser-schema-id: Language deliveryMode: &ref_10 type: string enum: - DELIVERY_MODE_UNSPECIFIED - STABLE - BALANCED - CREATIVE default: DELIVERY_MODE_UNSPECIFIED description: >- *Only supported by `inworld-tts-2`. The field is ignored on other models.* Controls how varied the output is. - `DELIVERY_MODE_UNSPECIFIED`: Defaults to `BALANCED` behavior. - `STABLE`: Optimizes for more consistent, predictable output. - `BALANCED`: Balanced between stability and diversity. - `CREATIVE`: Optimizes for increased emotional range and variation. x-parser-schema-id: DeliveryMode required: - voiceId - modelId x-parser-schema-id: contextId: type: string description: Optional context ID. If not provided, one will be generated x-parser-schema-id: required: - create examples: - create: voiceId: Dennis modelId: inworld-tts-2 bufferCharThreshold: 100 autoMode: true timestampType: WORD timestampTransportStrategy: ASYNC contextId: ctx-1 x-parser-schema-id: CreateContextPayload title: Create Context description: >- Create a new context with specified voice and configuration. A context is an independent conversation happening over the connection. The configurations for each context are completely separate – you can have different voice ids, models, output formats, etc. between contexts. *Note*: for each connection, 5 contexts is the max. If you don't need multiple contexts, you can omit the contextId in the message to use a single context connection. example: |- { "create": { "voiceId": "Dennis", "modelId": "inworld-tts-2", "bufferCharThreshold": 100, "autoMode": true, "timestampType": "WORD", "timestampTransportStrategy": "ASYNC" }, "contextId": "ctx-1" } bindings: [] extensions: - id: x-parser-unique-object-id value: createContext - &ref_16 id: sendText payload: - name: Send Text description: >- Send text to be synthesized for a specific context. You can only send up to 1000 characters in a single send_text request. Text can be buffered on the server or immediately flushed by including `flush_context` in the message. type: object properties: - name: send_text type: object required: true properties: - name: text type: string description: >- The text to synthesize. Maximum 1000 characters per send_text request. required: false - name: flush_context type: object description: >- Flush a context and start synthesis of all accumulated text required: false properties: [] - name: contextId type: string description: >- The target context for this message. Optional if only one context has been opened required: false headers: [] jsonPayloadSchema: type: object properties: send_text: type: object properties: text: type: string description: >- The text to synthesize. Maximum 1000 characters per send_text request. x-parser-schema-id: flush_context: type: object properties: {} description: Flush a context and start synthesis of all accumulated text x-parser-schema-id: required: - text x-parser-schema-id: contextId: type: string description: >- The target context for this message. Optional if only one context has been opened x-parser-schema-id: required: - send_text examples: - send_text: text: Hello, what a wonderful day to be a text-to-speech model! flush_context: {} contextId: ctx-1 x-parser-schema-id: SendTextPayload title: Send Text description: >- Send text to be synthesized for a specific context. You can only send up to 1000 characters in a single send_text request. Text can be buffered on the server or immediately flushed by including `flush_context` in the message. example: |- { "send_text": { "text": "Hello, what a wonderful day to be a text-to-speech model!", "flush_context": {} }, "contextId": "ctx-1" } bindings: [] extensions: - id: x-parser-unique-object-id value: sendText - &ref_17 id: flushContext payload: - name: Flush Context description: >- Flush a context and start synthesis of all accumulated text. Note that the buffer will automatically flush all text if the length of text is greater than 1000 characters, regardless of any other buffer settings. type: object properties: - name: flush_context type: object required: true properties: [] - name: contextId type: string description: >- The context to flush. Optional if only one context has been opened required: false headers: [] jsonPayloadSchema: type: object description: Flush a context and force a synthesis of all accumulated text properties: flush_context: type: object properties: {} x-parser-schema-id: contextId: type: string description: >- The context to flush. Optional if only one context has been opened x-parser-schema-id: required: - flush_context examples: - flush_context: {} contextId: ctx-1 x-parser-schema-id: FlushContext title: Flush Context description: >- Flush a context and start synthesis of all accumulated text. Note that the buffer will automatically flush all text if the length of text is greater than 1000 characters, regardless of any other buffer settings. example: |- { "flush_context": {}, "contextId": "ctx-1" } bindings: [] extensions: - id: x-parser-unique-object-id value: flushContext - &ref_18 id: closeContext payload: - name: Close Context description: >- Close an existing context and release all of its resources. Sending a close context message is equivalent to sending a flush message right before, so all text in the buffer will be synthesized before the context is closed. Note that the session will automatically be closed after 10 minutes of inactivity across any context. type: object properties: - name: close_context type: object required: true properties: [] - name: contextId type: string description: The context to close required: false headers: [] jsonPayloadSchema: type: object properties: close_context: type: object properties: {} x-parser-schema-id: contextId: type: string description: The context to close x-parser-schema-id: required: - close_context examples: - close_context: {} contextId: ctx-1 x-parser-schema-id: CloseContextPayload title: Close Context description: >- Close an existing context and release all of its resources. Sending a close context message is equivalent to sending a flush message right before, so all text in the buffer will be synthesized before the context is closed. Note that the session will automatically be closed after 10 minutes of inactivity across any context. example: |- { "close_context": {}, "contextId": "ctx-1" } bindings: [] extensions: - id: x-parser-unique-object-id value: closeContext bindings: [] extensions: &ref_12 - id: x-parser-unique-object-id value: ttsStream - &ref_14 id: receiveRequest title: Receive request type: send messages: - &ref_19 id: contextCreated payload: - name: Context Created description: Event sent when a new TTS context has been successfully created type: object properties: - name: result type: object required: false properties: - name: contextId type: string description: The ID of the created context required: false - name: contextCreated type: object required: false properties: - name: voiceId type: string description: The identifier of the voice to use for the synthesis required: false - name: modelId type: string description: >- The ID of the model to use for synthesizing speech. See [Models](../../../tts/tts-models) for available models. required: false - name: audioConfig type: object required: false properties: - name: audioEncoding type: string description: >- The desired output format of the synthesized audio. Defaults to `MP3`. - `LINEAR16`: Uncompressed 16-bit signed little-endian samples (Linear PCM). The WAV header is included in every audio chunk. - `MP3`: MP3 audio. - `OGG_OPUS`: Opus encoded audio wrapped in an ogg container. The result will be a file which can be played natively on Android, and in browsers (at least Chrome and Firefox). The quality of the encoding is considerably higher than MP3 while using approximately the same bitrate. - `ALAW`: ALAW encoded audio. 8-bit companded PCM. - `MULAW`: MULAW encoded audio. 8-bit companded PCM. - `PCM`: PCM audio. Uncompressed 16-bit signed little-endian samples with no WAV header. - `WAV`: WAV audio. Uncompressed 16-bit signed little-endian samples. The WAV header is included in the first audio chunk only. On each flush_completed response, the next audio chunk will also start with a header. enumValues: - AUDIO_ENCODING_UNSPECIFIED - LINEAR16 - MP3 - OGG_OPUS - ALAW - MULAW - PCM - WAV required: false - name: sampleRateHertz type: integer description: >- The synthesis sample rate (in hertz) for this audio. Accepts values within the range [8000, 48000]. When this is specified, if this is different from the voice's natural sample rate, then the audio will be converted to the desired sample rate (which might result in worse audio quality), unless the specified sample rate is not supported for the encoding chosen, in which case it will fail the request and return an error. The default is 48,000. required: false - name: bitRate type: integer description: >- Bits per second of the audio. Only for compressed audio formats (`MP3`, `OGG_OPUS`). The default is 128,000. required: false - name: speakingRate type: number description: >- Speaking rate/speed, in the range [0.5, 1.5]. The default is 1.0, which is the normal native speed supported by the specific voice. We recommend using values above 0.8 to ensure high quality. required: false - name: temperature type: number description: >- *Ignored on `inworld-tts-2`. Use [`deliveryMode`](#deliveryMode) instead.* Determines the degree of randomness when sampling audio tokens to generate the response. Defaults to 1.0. Accepts values between 0 (exclusive) and 2 (inclusive). Higher values will make the output more random and can lead to more expressive results. Lower values will make it more deterministic. If 0 is provided, the default value will be used. For the most stable results, we recommend using the default value. required: false - name: timestampType type: string description: >- Controls timestamp metadata returned with the audio. When enabled, the response includes timing arrays, which can be useful for word-highlighting, karaoke-style captions, and lipsync. - WORD: Output arrays under `timestampInfo.wordAlignment` (words, wordStartTimeSeconds, wordEndTimeSeconds). - CHARACTER: Output arrays under `timestampInfo.characterAlignment` (characters, characterStartTimeSeconds, characterEndTimeSeconds). - TIMESTAMP_TYPE_UNSPECIFIED: Do not compute alignment; timestamp arrays will be empty or omitted. **Latency note:** Alignment adds additional computation. Enabling alignment can increase latency. The timestamps reset per flush, either triggered manually or automatically by the server. When you receive a `flushCompleted` message, the timestamps for subsequent chunks will start from 0. enumValues: - TIMESTAMP_TYPE_UNSPECIFIED - WORD - CHARACTER required: false - name: maxBufferDelayMs type: integer description: >- If set, determines the maximum time in milliseconds to buffer before starting generation. The timer starts running when the first text in the buffer is received and resets when new text arrives. If not set or set to 0, there will be no time-based limit on the buffer. Instead, the user will need to flush the buffer or set the bufferCharThreshold to trigger audio generation. Note that both length and timeout based flushing can be used together - if both maxBufferDelayMs and bufferCharThreshold are set, the server will flush if either condition is met. required: false - name: bufferCharThreshold type: integer description: >- Defines the minimum number of characters in the buffer that would automatically trigger audio generation. This allows you to rely on automatic triggering instead of calling flush manually. If set to 0 or left undefined, the threshold defaults to 1000 to ensure stable behavior. Cannot be set to a value greater than 1000. Note that the buffer will automatically flush if more than 1000 characters are accumulated, regardless of this setting. required: false - name: applyTextNormalization type: string description: >- When enabled, text normalization automatically expands and standardizes things like numbers, dates, times, and abbreviations before converting them to speech. For example, Dr. Smith becomes Doctor Smith, and 3/10/25 is spoken as March tenth, twenty twenty-five. Turning this off may reduce latency, but the speech output will read the text exactly as written. Defaults to automatically deciding whether to apply text normalization. enumValues: - APPLY_TEXT_NORMALIZATION_UNSPECIFIED - 'ON' - 'OFF' required: false - name: autoMode type: boolean description: >- When enabled, the server will control flushing of buffered text to achieve minimal latency, while still maintaining high quality audio output. Recommended when texts are sent in full sentences/phrases. The default is false. required: false - name: timestampTransportStrategy type: string description: >- The transport strategy of timestamps info. - `TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED`: The service will automatically decide the transport strategy. - `SYNC`: Timestamps will be returned in the same message as the audio data. - `ASYNC`: Timestamps could return in trailing message after the audio data. Use this strategy to reduce latency of the first audio chunk with v1.5+ models. enumValues: - TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED - SYNC - ASYNC required: false - name: language type: string description: >- BCP-47 language tag (e.g., `en-US`, `fr-FR`, `ja-JP`) specifying the language that the given voice should speak the text in. If a localized voice prompt exists for the language, it will be used. When omitted, the original voice prompt will be used and the language will be auto-detected from the input text. If an invalid language code is provided, an error will be returned. See [Languages](../../../tts/capabilities/multilingual) for more details. required: false - name: deliveryMode type: string description: >- *Only supported by `inworld-tts-2`. The field is ignored on other models.* Controls how varied the output is. - `DELIVERY_MODE_UNSPECIFIED`: Defaults to `BALANCED` behavior. - `STABLE`: Optimizes for more consistent, predictable output. - `BALANCED`: Balanced between stability and diversity. - `CREATIVE`: Optimizes for increased emotional range and variation. enumValues: - DELIVERY_MODE_UNSPECIFIED - STABLE - BALANCED - CREATIVE required: false - name: status type: object description: >- Status information for gRPC responses, including any error details if applicable required: false properties: - name: code type: integer description: >- The status code, as specified by [gRPC status codes](https://grpc.io/docs/guides/status-codes/). required: false - name: message type: string description: A short description of the error required: false - name: details type: array description: Additional status or error details required: false headers: [] jsonPayloadSchema: type: object properties: result: type: object properties: contextId: type: string description: The ID of the created context x-parser-schema-id: contextCreated: type: object properties: voiceId: *ref_0 modelId: type: string description: >- The ID of the model to use for synthesizing speech. See [Models](../../../tts/tts-models) for available models. x-parser-schema-id: audioConfig: *ref_1 temperature: *ref_2 timestampType: *ref_3 maxBufferDelayMs: *ref_4 bufferCharThreshold: *ref_5 applyTextNormalization: *ref_6 autoMode: *ref_7 timestampTransportStrategy: *ref_8 language: *ref_9 deliveryMode: *ref_10 x-parser-schema-id: status: &ref_11 type: object properties: code: type: integer format: int32 description: >- The status code, as specified by [gRPC status codes](https://grpc.io/docs/guides/status-codes/). x-parser-schema-id: message: type: string description: A short description of the error x-parser-schema-id: details: type: array items: type: object x-parser-schema-id: description: Additional status or error details x-parser-schema-id: description: >- Status information for gRPC responses, including any error details if applicable x-parser-schema-id: Status x-parser-schema-id: examples: - result: contextId: ctx-1 contextCreated: voiceId: Dennis audioConfig: audioEncoding: LINEAR16 sampleRateHertz: 16000 modelId: inworld-tts-2 timestampType: WORD maxBufferDelayMs: 3000 autoMode: true timestampTransportStrategy: SYNC language: en-US deliveryMode: BALANCED status: code: 0 message: '' details: [] x-parser-schema-id: ContextCreatedPayload title: Context Created description: Event sent when a new TTS context has been successfully created example: |- { "result": { "contextId": "ctx-1", "contextCreated": { "voiceId": "Dennis", "audioConfig": { "audioEncoding": "LINEAR16", "sampleRateHertz": 16000 }, "modelId": "inworld-tts-2", "timestampType": "WORD", "maxBufferDelayMs": 3000, "autoMode": true, "timestampTransportStrategy": "SYNC", "language": "en-US", "deliveryMode": "BALANCED" }, "status": { "code": 0, "message": "", "details": [] } } } bindings: [] extensions: - id: x-parser-unique-object-id value: contextCreated - &ref_20 id: audioChunk payload: - name: Audio Chunk description: Audio data chunk containing synthesized speech type: object properties: - name: result type: object required: false properties: - name: contextId type: string description: The context this audio chunk belongs to required: false - name: audioChunk type: object required: false properties: - name: audioContent type: string description: Base64 encoded audio data required: false - name: usage type: object required: false properties: - name: processedCharactersCount type: integer description: >- Number of characters of the input text processed so far required: false - name: modelId type: string description: The model used for speech synthesis required: false - name: timestampInfo type: object description: >- Timestamp alignment information when alignment is enabled required: false properties: - name: wordAlignment type: object description: Word-level alignment when timestampType is WORD required: false properties: - name: words type: array description: Aligned words in order required: false - name: wordStartTimeSeconds type: array description: >- Start time for each word in seconds from the beginning of the audio required: false - name: wordEndTimeSeconds type: array description: >- End time for each word in seconds from the beginning of the audio required: false - name: characterAlignment type: object description: >- Character-level alignment when timestampType is CHARACTER required: false properties: - name: characters type: array description: >- Aligned characters (including punctuation and spaces) in order required: false - name: characterStartTimeSeconds type: array description: >- Start time for each character in seconds from the beginning of the audio required: false - name: characterEndTimeSeconds type: array description: >- End time for each character in seconds from the beginning of the audio required: false - name: status type: object description: >- Status information for gRPC responses, including any error details if applicable required: false properties: - name: code type: integer description: >- The status code, as specified by [gRPC status codes](https://grpc.io/docs/guides/status-codes/). required: false - name: message type: string description: A short description of the error required: false - name: details type: array description: Additional status or error details required: false headers: [] jsonPayloadSchema: type: object properties: result: type: object properties: contextId: type: string description: The context this audio chunk belongs to x-parser-schema-id: audioChunk: type: object properties: audioContent: type: string format: byte description: Base64 encoded audio data x-parser-schema-id: usage: type: object properties: processedCharactersCount: type: integer format: int32 description: >- Number of characters of the input text processed so far x-parser-schema-id: modelId: type: string description: The model used for speech synthesis x-parser-schema-id: x-parser-schema-id: timestampInfo: type: object properties: wordAlignment: type: object properties: words: type: array items: type: string x-parser-schema-id: description: Aligned words in order x-parser-schema-id: wordStartTimeSeconds: type: array items: type: number x-parser-schema-id: description: >- Start time for each word in seconds from the beginning of the audio x-parser-schema-id: wordEndTimeSeconds: type: array items: type: number x-parser-schema-id: description: >- End time for each word in seconds from the beginning of the audio x-parser-schema-id: description: Word-level alignment when timestampType is WORD x-parser-schema-id: WordAlignment characterAlignment: type: object properties: characters: type: array items: type: string x-parser-schema-id: description: >- Aligned characters (including punctuation and spaces) in order x-parser-schema-id: characterStartTimeSeconds: type: array items: type: number x-parser-schema-id: description: >- Start time for each character in seconds from the beginning of the audio x-parser-schema-id: characterEndTimeSeconds: type: array items: type: number x-parser-schema-id: description: >- End time for each character in seconds from the beginning of the audio x-parser-schema-id: description: >- Character-level alignment when timestampType is CHARACTER x-parser-schema-id: CharacterAlignment description: >- Timestamp alignment information when alignment is enabled x-parser-schema-id: TimestampInfo status: *ref_11 x-parser-schema-id: x-parser-schema-id: examples: - result: contextId: ctx-1 audioChunk: audioContent: UklGRgSYAABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YeCX= usage: processedCharactersCount: 79 modelId: inworld-tts-2 timestampInfo: wordAlignment: words: - Hello, - what - a - wonderful - day - to - be - a - text-to-speech - model. wordStartTimeSeconds: - 0.031 - 0.375 - 0.901 - 1.002 - 1.386 - 1.548 - 1.649 - 1.771 - 1.852 - 2.58 wordEndTimeSeconds: - 0.355 - 0.86 - 0.921 - 1.326 - 1.528 - 1.609 - 1.71 - 1.791 - 2.539 - 2.802 status: code: 0 message: '' details: [] x-parser-schema-id: AudioChunkPayload title: Audio Chunk description: Audio data chunk containing synthesized speech example: |- { "result": { "contextId": "ctx-1", "audioChunk": { "audioContent": "UklGRgSYAABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YeCX=", "usage": { "processedCharactersCount": 79, "modelId": "inworld-tts-2" }, "timestampInfo": { "wordAlignment": { "words": [ "Hello,", "what", "a", "wonderful", "day", "to", "be", "a", "text-to-speech", "model." ], "wordStartTimeSeconds": [ 0.031, 0.375, 0.901, 1.002, 1.386, 1.548, 1.649, 1.771, 1.852, 2.58 ], "wordEndTimeSeconds": [ 0.355, 0.86, 0.921, 1.326, 1.528, 1.609, 1.71, 1.791, 2.539, 2.802 ] } }, "status": { "code": 0, "message": "", "details": [] } } } } bindings: [] extensions: - id: x-parser-unique-object-id value: audioChunk - &ref_21 id: contextClosed payload: - name: Context Closed description: Event sent when a context has been closed type: object properties: - name: result type: object required: false properties: - name: contextId type: string description: The ID of the closed context required: false - name: contextClosed type: object required: false properties: [] - name: status type: object description: >- Status information for gRPC responses, including any error details if applicable required: false properties: - name: code type: integer description: >- The status code, as specified by [gRPC status codes](https://grpc.io/docs/guides/status-codes/). required: false - name: message type: string description: A short description of the error required: false - name: details type: array description: Additional status or error details required: false headers: [] jsonPayloadSchema: type: object properties: result: type: object properties: contextId: type: string description: The ID of the closed context x-parser-schema-id: contextClosed: type: object properties: {} x-parser-schema-id: status: *ref_11 x-parser-schema-id: examples: - result: contextId: ctx-1 contextClosed: {} status: code: 0 message: '' details: [] x-parser-schema-id: ContextClosedPayload title: Context Closed description: Event sent when a context has been closed example: |- { "result": { "contextId": "ctx-1", "contextClosed": {}, "status": { "code": 0, "message": "", "details": [] } } } bindings: [] extensions: - id: x-parser-unique-object-id value: contextClosed - &ref_22 id: flushCompleted payload: - name: Flush Completed description: >- Event sent when speech synthesis for a flush of text is completed. Some websocket use cases require an indicator that speech synthesis for a flush of text is completed. To facilitate this, we've included an empty "flushCompleted":{} event at the end of speech synthesis for each flush. Note that the implementation currently assumes that flushes execute sequentially, so the first flushCompleted event would correspond to the first flush call made on the client side. type: object properties: - name: result type: object required: false properties: - name: contextId type: string description: The ID of the context for which the flush completed required: false - name: flushCompleted type: object required: false properties: [] - name: status type: object description: >- Status information for gRPC responses, including any error details if applicable required: false properties: - name: code type: integer description: >- The status code, as specified by [gRPC status codes](https://grpc.io/docs/guides/status-codes/). required: false - name: message type: string description: A short description of the error required: false - name: details type: array description: Additional status or error details required: false headers: [] jsonPayloadSchema: type: object properties: result: type: object properties: contextId: type: string description: The ID of the context for which the flush completed x-parser-schema-id: flushCompleted: type: object properties: {} x-parser-schema-id: status: *ref_11 x-parser-schema-id: examples: - result: contextId: ctx-1 flushCompleted: {} status: code: 0 message: '' details: [] x-parser-schema-id: FlushCompletedPayload title: Flush Completed description: >- Event sent when speech synthesis for a flush of text is completed. Some websocket use cases require an indicator that speech synthesis for a flush of text is completed. To facilitate this, we've included an empty "flushCompleted":{} event at the end of speech synthesis for each flush. Note that the implementation currently assumes that flushes execute sequentially, so the first flushCompleted event would correspond to the first flush call made on the client side. example: |- { "result": { "contextId": "ctx-1", "flushCompleted": {}, "status": { "code": 0, "message": "", "details": [] } } } bindings: [] extensions: - id: x-parser-unique-object-id value: flushCompleted bindings: [] extensions: *ref_12 sendOperations: - *ref_13 receiveOperations: - *ref_14 sendMessages: - *ref_15 - *ref_16 - *ref_17 - *ref_18 receiveMessages: - *ref_19 - *ref_20 - *ref_21 - *ref_22 extensions: - id: x-parser-unique-object-id value: ttsStream securitySchemes: - id: auth name: authorization type: httpApiKey description: >- Your [authentication](../../../api-reference/introduction) credentials. For Basic authentication, please populate `Basic $INWORLD_API_KEY` in: query extensions: [] ````