> ## Documentation Index > Fetch the complete documentation index at: https://docs.slng.ai/llms.txt > Use this file to discover all available pages before exploring further. # Cartesia Sonic 3 > Stream low-latency speech synthesis from Cartesia Sonic 3 over WebSocket with a multilingual catalog of voices and context-aware controls. **Hosting: routed through SLNG (Cartesia).** This model runs on Cartesia's servers; SLNG proxies each request through the nearest regional gateway. Available in Netherlands, Germany, United States (East, West), Australia, Brazil, United Kingdom, Indonesia, Israel, India, Japan, Singapore, South Africa. Connect to your region's gateway WebSocket endpoint: `wss://eu-north.api.slng.ai`, `wss://eu-west.api.slng.ai`, `wss://us-east.api.slng.ai`, `wss://us-west.api.slng.ai`, `wss://au.api.slng.ai`, `wss://br.api.slng.ai`, `wss://gb.api.slng.ai`, `wss://id.api.slng.ai`, `wss://il.api.slng.ai`, `wss://in.api.slng.ai`, `wss://jp.api.slng.ai`, `wss://sg.api.slng.ai`, `wss://za.api.slng.ai`. ## AsyncAPI ````yaml api-reference/asyncapi/tts-cartesia.asyncapi.yaml /v1/tts/cartesia/sonic:3 id: /v1/tts/cartesia/sonic:3 title: Cartesia Sonic 3 description: >- Stream low-latency speech synthesis from Cartesia Sonic 3 over WebSocket with a multilingual catalog of voices and context-aware controls. servers: - id: production-eu-north protocol: wss host: eu-north.api.slng.ai bindings: [] variables: [] - id: production-eu-west protocol: wss host: eu-west.api.slng.ai bindings: [] variables: [] - id: production-us-east protocol: wss host: us-east.api.slng.ai bindings: [] variables: [] - id: production-us-west protocol: wss host: us-west.api.slng.ai bindings: [] variables: [] - id: production-au protocol: wss host: au.api.slng.ai bindings: [] variables: [] - id: production-br protocol: wss host: br.api.slng.ai bindings: [] variables: [] - id: production-gb protocol: wss host: gb.api.slng.ai bindings: [] variables: [] - id: production-id protocol: wss host: id.api.slng.ai bindings: [] variables: [] - id: production-il protocol: wss host: il.api.slng.ai bindings: [] variables: [] - id: production-in protocol: wss host: in.api.slng.ai bindings: [] variables: [] - id: production-jp protocol: wss host: jp.api.slng.ai bindings: [] variables: [] - id: production-sg protocol: wss host: sg.api.slng.ai bindings: [] variables: [] - id: production-za protocol: wss host: za.api.slng.ai bindings: [] variables: [] address: /v1/tts/cartesia/sonic:3 parameters: [] bindings: - protocol: ws version: latest value: method: GET schemaProperties: - name: method type: string description: GET required: false operations: - &ref_2 id: ttsCartesiaSonic3ReceiveInit title: Tts cartesia sonic3 receive init description: Initialize session on Cartesia Sonic 3 type: receive messages: - &ref_11 id: CartesiaSonicTtsInitRequest contentType: application/json payload: - name: Init Request (Cartesia Sonic 3) description: >- Initialize a Cartesia Sonic 3 TTS session with provider-specific configuration. type: object properties: - name: type type: string description: init required: true - name: model_id type: string description: Cartesia Sonic model identifier. enumValues: - sonic-3 required: false - name: voice type: string description: Cartesia voice identifier. required: false - name: language type: string description: Optional language override for supported Sonic 3 voices. required: false - name: encoding type: string description: Output audio encoding format. enumValues: - linear16 - mulaw - ulaw - alaw required: false - name: sample_rate type: integer description: Output audio sample rate in Hz. enumValues: - 8000 - 16000 - 24000 - 44100 required: false - name: context_id type: string description: >- Optional synthesis context identifier. If omitted, the gateway generates one per WebSocket session. required: false headers: [] jsonPayloadSchema: type: object description: >- Initialize a Cartesia Sonic 3 TTS session with voice, language, and audio format options. Fields set here are applied for the session unless overridden per text message. required: - type properties: type: type: string const: init default: init x-parser-schema-id: model_id: type: string description: Cartesia Sonic model identifier. enum: - sonic-3 default: sonic-3 x-parser-schema-id: voice: type: string description: Cartesia voice identifier. default: f786b574-daa5-4673-aa0c-cbe3e8534c02 x-parser-schema-id: language: type: string description: Optional language override for supported Sonic 3 voices. x-parser-schema-id: encoding: type: string description: Output audio encoding format. enum: - linear16 - mulaw - ulaw - alaw default: linear16 x-parser-schema-id: sample_rate: type: integer description: Output audio sample rate in Hz. enum: - 8000 - 16000 - 24000 - 44100 default: 24000 x-parser-schema-id: context_id: type: string description: >- Optional synthesis context identifier. If omitted, the gateway generates one per WebSocket session. x-parser-schema-id: x-parser-schema-id: title: Init Request (Cartesia Sonic 3) description: >- Initialize a Cartesia Sonic 3 TTS session with provider-specific configuration. example: |- { "type": "init", "model_id": "sonic-3", "voice": "f786b574-daa5-4673-aa0c-cbe3e8534c02", "encoding": "linear16", "sample_rate": 24000 } bindings: [] extensions: - id: x-parser-unique-object-id value: CartesiaSonicTtsInitRequest bindings: [] extensions: &ref_0 - id: x-parser-unique-object-id value: /v1/tts/cartesia/sonic:3 - &ref_3 id: ttsCartesiaSonic3ReceiveText title: Tts cartesia sonic3 receive text description: Send text to Cartesia Sonic 3 type: receive messages: - &ref_12 id: CartesiaSonicTtsTextMessage contentType: application/json payload: - name: Text Message (Cartesia Sonic 3) description: >- Send text to synthesize on Cartesia Sonic 3 with optional generation controls. type: object properties: - name: type type: string description: text required: true - name: text type: string description: Text to synthesize. required: true - name: voice type: string description: Override the Cartesia voice identifier for this utterance. required: false - name: generation_config type: object description: >- Optional Sonic 3 generation controls such as speed, volume, and emotion. required: false properties: - name: speed type: number description: Sonic 3 speed multiplier guidance. required: false - name: volume type: number description: Sonic 3 volume multiplier guidance. required: false - name: emotion type: string description: Sonic 3 emotional guidance. required: false - name: speed type: string description: >- Deprecated Cartesia speed override. Prefer `generation_config.speed` for Sonic 3. Ignored when `generation_config` is provided. enumValues: - slow - normal - fast required: false - name: continue type: boolean description: >- Whether more text inputs may follow on the same `context_id`. When `true`, Cartesia keeps the context open and waits. When `false` or omitted, this is the final input — Cartesia emits the trailing audio and then `audio_end`. required: false - name: max_buffer_delay_ms type: integer description: >- Maximum time, in milliseconds, that Cartesia buffers incoming text before starting generation. Use it to group several quick `text` inputs into a single synthesis pass — the longer the delay, the more text Cartesia can combine. Omit or set `0` to generate as soon as the first text input arrives. required: false - name: add_timestamps type: boolean description: >- Whether Cartesia should emit timestamp events for this generation. required: false - name: flush type: boolean description: >- Marks a segment boundary within the active context. Not required to receive audio — Cartesia begins generating as soon as text arrives. Sending `flush: true` increments the `flush_id` that Cartesia includes on subsequent `audio_chunk` frames so you can correlate audio back to the text segment that produced it. Does not finalize the context; complete generation by sending the final input with `continue: false`. required: false - name: pronunciation_dict_id type: string description: >- Optional pronunciation dictionary applied to this generation only. required: false - name: context_id type: string description: >- Optional synthesis context identifier. If omitted, the gateway uses the active session context_id. required: false headers: [] jsonPayloadSchema: type: object description: >- Send text to synthesize into audio output using Cartesia Sonic 3. Supports per-utterance generation controls including speed, volume, emotion, and manual flushing. required: - type - text properties: type: type: string const: text x-parser-schema-id: text: type: string minLength: 1 description: Text to synthesize. x-parser-schema-id: voice: type: string description: Override the Cartesia voice identifier for this utterance. x-parser-schema-id: generation_config: type: object description: >- Optional Sonic 3 generation controls such as speed, volume, and emotion. properties: speed: type: number minimum: 0.6 maximum: 1.5 description: Sonic 3 speed multiplier guidance. x-parser-schema-id: volume: type: number minimum: 0.5 maximum: 2 description: Sonic 3 volume multiplier guidance. x-parser-schema-id: emotion: type: string description: Sonic 3 emotional guidance. x-parser-schema-id: x-parser-schema-id: speed: type: string description: >- Deprecated Cartesia speed override. Prefer `generation_config.speed` for Sonic 3. Ignored when `generation_config` is provided. enum: - slow - normal - fast x-parser-schema-id: continue: type: boolean description: >- Whether more text inputs may follow on the same `context_id`. When `true`, Cartesia keeps the context open and waits. When `false` or omitted, this is the final input — Cartesia emits the trailing audio and then `audio_end`. x-parser-schema-id: max_buffer_delay_ms: type: integer minimum: 0 maximum: 5000 description: >- Maximum time, in milliseconds, that Cartesia buffers incoming text before starting generation. Use it to group several quick `text` inputs into a single synthesis pass — the longer the delay, the more text Cartesia can combine. Omit or set `0` to generate as soon as the first text input arrives. x-parser-schema-id: add_timestamps: type: boolean description: >- Whether Cartesia should emit timestamp events for this generation. x-parser-schema-id: flush: type: boolean description: >- Marks a segment boundary within the active context. Not required to receive audio — Cartesia begins generating as soon as text arrives. Sending `flush: true` increments the `flush_id` that Cartesia includes on subsequent `audio_chunk` frames so you can correlate audio back to the text segment that produced it. Does not finalize the context; complete generation by sending the final input with `continue: false`. x-parser-schema-id: pronunciation_dict_id: type: string description: >- Optional pronunciation dictionary applied to this generation only. x-parser-schema-id: context_id: type: string description: >- Optional synthesis context identifier. If omitted, the gateway uses the active session context_id. x-parser-schema-id: x-parser-schema-id: title: Text Message (Cartesia Sonic 3) description: >- Send text to synthesize on Cartesia Sonic 3 with optional generation controls. example: |- { "type": "text", "text": "Hello, this is Cartesia Sonic 3." } bindings: [] extensions: - id: x-parser-unique-object-id value: CartesiaSonicTtsTextMessage bindings: [] extensions: *ref_0 - &ref_4 id: ttsCartesiaSonic3ReceiveFlush title: Tts cartesia sonic3 receive flush description: Flush buffered output on Cartesia Sonic 3 type: receive messages: - &ref_13 id: TtsFlushMessage contentType: application/json payload: - name: Flush Message description: >- Standalone segment boundary on the active Cartesia Sonic 3 context. type: object properties: - name: type type: string description: flush required: true headers: [] jsonPayloadSchema: type: object description: >- Standalone segment boundary on the active context. Has the same effect as setting `flush: true` on a `text` input: it increments `flush_id` and does not finalize the context. Complete generation by sending the final `text` input with `continue: false`. required: - type properties: type: type: string const: flush x-parser-schema-id: x-parser-schema-id: title: Flush Message description: Standalone segment boundary on the active Cartesia Sonic 3 context. example: |- { "type": "flush" } bindings: [] extensions: - id: x-parser-unique-object-id value: TtsFlushMessage bindings: [] extensions: *ref_0 - &ref_5 id: ttsCartesiaSonic3ReceiveCancel title: Tts cartesia sonic3 receive cancel description: Cancel in-progress generation on Cartesia Sonic 3 type: receive messages: - &ref_14 id: TtsCancelMessage contentType: application/json payload: - name: Cancel Message description: Cancel an in-progress Cartesia Sonic 3 synthesis context. type: object properties: - name: type type: string description: cancel required: true - name: context_id type: string description: >- Optional context identifier to cancel. If omitted, the gateway cancels the active session context_id. required: false headers: [] jsonPayloadSchema: type: object description: >- Cancel an in-progress Cartesia synthesis context. The gateway forwards a native `{ context_id, cancel: true }` frame to Cartesia. When `context_id` is omitted, the active session context is cancelled. Cartesia recommends starting a new `context_id` for the next turn after a cancel. required: - type properties: type: type: string const: cancel x-parser-schema-id: context_id: type: string description: >- Optional context identifier to cancel. If omitted, the gateway cancels the active session context_id. x-parser-schema-id: x-parser-schema-id: title: Cancel Message description: Cancel an in-progress Cartesia Sonic 3 synthesis context. example: |- { "type": "cancel" } bindings: [] extensions: - id: x-parser-unique-object-id value: TtsCancelMessage bindings: [] extensions: *ref_0 - &ref_6 id: ttsCartesiaSonic3ReceiveClose title: Tts cartesia sonic3 receive close description: Close session on Cartesia Sonic 3 type: receive messages: - &ref_15 id: TtsCloseMessage contentType: application/json payload: - name: Close Message description: Close the session and stop any further audio. type: object properties: - name: type type: string description: close required: true headers: [] jsonPayloadSchema: type: object description: Close the session and stop any further audio. required: - type properties: type: type: string const: close x-parser-schema-id: x-parser-schema-id: title: Close Message description: Close the session and stop any further audio. example: |- { "type": "close" } bindings: [] extensions: - id: x-parser-unique-object-id value: TtsCloseMessage bindings: [] extensions: *ref_0 - &ref_7 id: ttsCartesiaSonic3ReceiveKeepalive title: Tts cartesia sonic3 receive keepalive description: Send keepalive on Cartesia Sonic 3 type: receive messages: - &ref_16 id: KeepaliveMessage contentType: application/json payload: - name: Keepalive Message description: Keep the WebSocket connection alive during silence. type: object properties: - name: type type: string description: keepalive required: true headers: [] jsonPayloadSchema: type: object description: >- Prevents the connection from being closed due to inactivity. Send periodically during silence to maintain the session. required: - type properties: type: type: string const: keepalive x-parser-schema-id: x-parser-schema-id: title: Keepalive Message description: Keep the WebSocket connection alive during silence. example: |- { "type": "keepalive" } bindings: [] extensions: - id: x-parser-unique-object-id value: KeepaliveMessage bindings: [] extensions: *ref_0 - &ref_8 id: ttsCartesiaSonic3SendAudioChunk title: Tts cartesia sonic3 send audio chunk description: Receive audio chunk from Cartesia Sonic 3 type: send messages: - &ref_17 id: CartesiaSonicTtsAudioChunk contentType: application/json payload: - name: Audio Chunk (Cartesia Sonic 3) description: Streaming audio chunk from Cartesia Sonic 3. type: object properties: - name: type type: string description: audio_chunk required: true - name: data type: string description: >- Base64-encoded audio chunk matching the requested encoding and sample rate. required: true - name: done type: boolean description: True when this is the final chunk of the active context. required: false - name: status_code type: integer description: Upstream HTTP-equivalent status code from Cartesia. required: false - name: step_time type: number description: Cartesia processing time for this chunk, in milliseconds. required: false - name: flush_id type: integer description: >- Segment identifier within the active context. Starts at `0` and increments by `1` each time a `text` input is sent with `flush: true`, or whenever a standalone flush message is issued. Use it to correlate audio chunks back to the text segment that produced them. required: false - name: context_id type: string description: Synthesis context identifier this chunk belongs to. required: false headers: [] jsonPayloadSchema: type: object description: >- Streaming audio chunk emitted by Cartesia Sonic 3. The gateway forwards each upstream `chunk` message as an `audio_chunk`, preserving Cartesia's native fields. required: - type - data properties: type: type: string const: audio_chunk x-parser-schema-id: data: type: string format: byte description: >- Base64-encoded audio chunk matching the requested encoding and sample rate. x-parser-schema-id: done: type: boolean description: True when this is the final chunk of the active context. x-parser-schema-id: status_code: type: integer description: Upstream HTTP-equivalent status code from Cartesia. x-parser-schema-id: step_time: type: number description: Cartesia processing time for this chunk, in milliseconds. x-parser-schema-id: flush_id: type: integer description: >- Segment identifier within the active context. Starts at `0` and increments by `1` each time a `text` input is sent with `flush: true`, or whenever a standalone flush message is issued. Use it to correlate audio chunks back to the text segment that produced them. x-parser-schema-id: context_id: type: string description: Synthesis context identifier this chunk belongs to. x-parser-schema-id: x-parser-schema-id: title: Audio Chunk (Cartesia Sonic 3) description: Streaming audio chunk from Cartesia Sonic 3. example: |- { "type": "audio_chunk", "context_id": "9baf38d7-799e-40f2-88d1-daeacbdb3910", "status_code": 206, "done": false, "data": "AAF+/dD7n/wj/rD/owGoBEQHBgjTBcoDjAITAk0BtwF0BFUI5wmlB7cEAgO3ArIBUwB0AJwDDQbQBacCEgDA/vD96vxr/H3+hwAqASr/Z/3C/BEAEQAUAA==", "step_time": 67.49622, "flush_id": 0 } bindings: [] extensions: - id: x-parser-unique-object-id value: CartesiaSonicTtsAudioChunk bindings: [] extensions: *ref_0 - &ref_9 id: ttsCartesiaSonic3SendAudioEnd title: Tts cartesia sonic3 send audio end description: Receive audio end from Cartesia Sonic 3 type: send messages: - &ref_18 id: TtsAudioEnd contentType: application/json payload: - name: Audio End description: Signals the end of audio generation. type: object properties: - name: type type: string description: audio_end required: true - name: done type: boolean description: Always `true` on Cartesia generation-complete frames. required: false - name: status_code type: integer description: Upstream HTTP-equivalent status code from Cartesia. required: false - name: context_id type: string description: Synthesis context identifier that finished generation. required: false headers: [] jsonPayloadSchema: type: object description: >- Signals the end of audio generation for the active context. The gateway maps the upstream Cartesia `done` message to `audio_end`, preserving Cartesia's native fields. required: - type properties: type: type: string const: audio_end x-parser-schema-id: done: type: boolean description: Always `true` on Cartesia generation-complete frames. x-parser-schema-id: status_code: type: integer description: Upstream HTTP-equivalent status code from Cartesia. x-parser-schema-id: context_id: type: string description: Synthesis context identifier that finished generation. x-parser-schema-id: x-parser-schema-id: title: Audio End description: Signals the end of audio generation. example: |- { "type": "audio_end", "context_id": "9baf38d7-799e-40f2-88d1-daeacbdb3910", "status_code": 200, "done": true } bindings: [] extensions: - id: x-parser-unique-object-id value: TtsAudioEnd bindings: [] extensions: *ref_0 - &ref_10 id: ttsCartesiaSonic3SendError title: Tts cartesia sonic3 send error description: Receive error from Cartesia Sonic 3 type: send messages: - &ref_19 id: TtsErrorResponse contentType: application/json payload: - name: Error Response description: Indicates an error occurred during synthesis. type: object properties: - name: type type: string description: error required: true - name: message type: string description: Human-readable error description. required: true - name: title type: string description: Short error title from the upstream provider. required: false - name: status_code type: integer description: Upstream HTTP-equivalent status code. required: false - name: error_code type: &ref_1 - string - 'null' description: Provider-specific error code, when available. required: false - name: done type: boolean description: >- True when the upstream marked the context as terminated by this error. required: false - name: request_id type: string description: Cartesia request identifier, useful when contacting support. required: false - name: doc_url type: string description: >- Cartesia documentation link for this error class, when provided. required: false - name: context_id type: string description: Synthesis context identifier associated with the error. required: false headers: [] jsonPayloadSchema: type: object description: >- Indicates an error occurred during synthesis. The gateway surfaces upstream Cartesia error messages with the original status code, provider-specific error code, and supporting metadata. required: - type - message properties: type: type: string const: error x-parser-schema-id: message: type: string description: Human-readable error description. x-parser-schema-id: title: type: string description: Short error title from the upstream provider. x-parser-schema-id: status_code: type: integer description: Upstream HTTP-equivalent status code. x-parser-schema-id: error_code: type: *ref_1 description: Provider-specific error code, when available. x-parser-schema-id: done: type: boolean description: >- True when the upstream marked the context as terminated by this error. x-parser-schema-id: request_id: type: string description: Cartesia request identifier, useful when contacting support. x-parser-schema-id: doc_url: type: string description: Cartesia documentation link for this error class, when provided. x-parser-schema-id: context_id: type: string description: Synthesis context identifier associated with the error. x-parser-schema-id: x-parser-schema-id: title: Error Response description: Indicates an error occurred during synthesis. example: |- { "type": "error", "message": "Provider returned an unexpected error", "status_code": 500, "error_code": "provider_error", "done": true, "request_id": "req-9b1c2f", "doc_url": "https://docs.cartesia.ai/errors/provider_error", "context_id": "demo-context-123" } bindings: [] extensions: - id: x-parser-unique-object-id value: TtsErrorResponse bindings: [] extensions: *ref_0 sendOperations: - *ref_2 - *ref_3 - *ref_4 - *ref_5 - *ref_6 - *ref_7 receiveOperations: - *ref_8 - *ref_9 - *ref_10 sendMessages: - *ref_11 - *ref_12 - *ref_13 - *ref_14 - *ref_15 - *ref_16 receiveMessages: - *ref_17 - *ref_18 - *ref_19 extensions: - id: x-parser-unique-object-id value: /v1/tts/cartesia/sonic:3 securitySchemes: - id: bearer name: bearer type: http description: > API key issued by SLNG. Pass as `Authorization: Bearer ` in the WebSocket upgrade request headers. scheme: bearer extensions: [] ````