> ## Documentation Index > Fetch the complete documentation index at: https://docs.chariot.in/llms.txt > Use this file to discover all available pages before exploring further. # Multi-context TTS WebSocket > Beta. Several independent speech streams (contexts) over one connection. Every message names its `context_id`. The first `input.text` for a new `context_id` opens the context; `input.flush` generates its buffered text as the next segment; `context.close` finishes it and `context.cancel` stops it immediately. Each context ends with one `context.done`, and nothing is sent for it afterwards. Up to 5 contexts can be open at once. All messages are JSON text frames. Audio arrives in `audio.chunk` messages as base64-encoded raw PCM, signed 16-bit little-endian, mono. Authenticate with the `chariotai-api-key` header, or with the `api_key` query parameter when your client cannot set handshake headers. The connection closes with code `1000` after `session.close`, `1008` on an auth or parameter problem or the idle timeout, `1011` on an internal error, and `4402` when the workspace has insufficient credits. Every other error affects a single message or context and leaves the connection open. ## AsyncAPI ````yaml api-reference/asyncapi.yaml ttsMultiStream id: ttsMultiStream title: Multi-context text to speech stream (beta) description: > Beta. Several independent speech streams (contexts) over one connection. Every message names its `context_id`. The first `input.text` for a new `context_id` opens the context; `input.flush` generates its buffered text as the next segment; `context.close` finishes it and `context.cancel` stops it immediately. Each context ends with one `context.done`, and nothing is sent for it afterwards. Up to 5 contexts can be open at once. All messages are JSON text frames. Audio arrives in `audio.chunk` messages as base64-encoded raw PCM, signed 16-bit little-endian, mono. Authenticate with the `chariotai-api-key` header, or with the `api_key` query parameter when your client cannot set handshake headers. The connection closes with code `1000` after `session.close`, `1008` on an auth or parameter problem or the idle timeout, `1011` on an internal error, and `4402` when the workspace has insufficient credits. Every other error affects a single message or context and leaves the connection open. servers: - id: production protocol: wss host: api.chariot.in bindings: [] variables: [] address: /v1/tts/ws-multi parameters: [] bindings: - protocol: ws version: 0.1.0 value: method: GET query: type: object properties: voice_id: type: string format: uuid description: > Default voice for contexts that do not set `voice_id` on their first message. Optional. Get one from `GET /v1/voices`. examples: &ref_0 - bac7d666-094d-4698-91fa-741d60fce662 model_type: type: string enum: - chariot-v0 - chariot-multilingual-v1 default: chariot-v0 examples: &ref_1 - chariot-multilingual-v1 description: > Model for the whole connection. Defaults to `chariot-v0`. Every voice used must support it. speed: type: number default: 1 minimum: 0.5 maximum: 2 examples: &ref_2 - 1 description: > Default speaking rate for contexts. Choose a value from the voice model's `speeds` list. `chariot-v0` supports only 1.0. optimize_streaming_latency: type: integer default: 0 minimum: 0 maximum: 4 examples: &ref_3 - 0 description: > Latency-optimization level for the whole connection. Defaults to 0, which applies all quality passes. idle_timeout: type: number default: 60 exclusiveMinimum: 0 maximum: 300 examples: &ref_4 - 60 description: > Seconds the server waits for any client message before closing the connection as idle. Must be greater than 0 and at most 300. enable_logging: type: boolean default: false examples: &ref_5 - false description: > When true, the audio for each generated segment is stored and you can retrieve it from your TTS history. Defaults to false. api_key: type: string description: > Your API key. Prefer the `chariotai-api-key` header where your client supports it. schemaProperties: - name: method type: string description: GET required: false - name: query type: object required: false properties: - name: voice_id type: string description: > Default voice for contexts that do not set `voice_id` on their first message. Optional. Get one from `GET /v1/voices`. examples: *ref_0 required: false - name: model_type type: string description: > Model for the whole connection. Defaults to `chariot-v0`. Every voice used must support it. enumValues: - chariot-v0 - chariot-multilingual-v1 examples: *ref_1 required: false - name: speed type: number description: > Default speaking rate for contexts. Choose a value from the voice model's `speeds` list. `chariot-v0` supports only 1.0. examples: *ref_2 required: false - name: optimize_streaming_latency type: integer description: > Latency-optimization level for the whole connection. Defaults to 0, which applies all quality passes. examples: *ref_3 required: false - name: idle_timeout type: number description: > Seconds the server waits for any client message before closing the connection as idle. Must be greater than 0 and at most 300. examples: *ref_4 required: false - name: enable_logging type: boolean description: > When true, the audio for each generated segment is stored and you can retrieve it from your TTS history. Defaults to false. examples: *ref_5 required: false - name: api_key type: string description: > Your API key. Prefer the `chariotai-api-key` header where your client supports it. required: false operations: - &ref_14 id: sendToChariotMulti title: Client to server description: Messages you send to Chariot. type: receive messages: - &ref_16 id: inputText payload: - name: Append text description: > Append text to a context's buffer. The first input.text for a context_id opens the context. Does not generate audio on its own. type: object properties: - name: type type: string description: input.text required: true - name: context_id type: string description: The context this message is for. examples: &ref_6 - turn-1 required: false - name: text type: string description: > Text to append. Each fragment must fit 500 characters for `chariot-v0` or 1,000 for `chariot-multilingual-v1`. If the context's unflushed text would exceed that limit, the buffered text is generated as a segment first, so no text is dropped. An empty string keeps the context alive without generating anything. examples: &ref_7 - 'Your order shipped on Monday. ' required: true - name: voice_id type: string description: > Voice for this context. Only read on the context's first message; defaults to the connection's `voice_id`. required: false - name: speed type: number description: > Speaking rate for this context. Only read on the context's first message; defaults to the connection's `speed`. required: false headers: [] jsonPayloadSchema: type: object required: - type - text properties: type: type: string const: input.text x-parser-schema-id: context_id: type: string pattern: ^[A-Za-z0-9_.:-]{1,128}$ default: default description: The context this message is for. examples: *ref_6 x-parser-schema-id: text: type: string description: > Text to append. Each fragment must fit 500 characters for `chariot-v0` or 1,000 for `chariot-multilingual-v1`. If the context's unflushed text would exceed that limit, the buffered text is generated as a segment first, so no text is dropped. An empty string keeps the context alive without generating anything. examples: *ref_7 x-parser-schema-id: voice_id: type: string format: uuid description: > Voice for this context. Only read on the context's first message; defaults to the connection's `voice_id`. x-parser-schema-id: speed: type: number description: > Speaking rate for this context. Only read on the context's first message; defaults to the connection's `speed`. x-parser-schema-id: x-parser-schema-id: title: Append text description: > Append text to a context's buffer. The first input.text for a context_id opens the context. Does not generate audio on its own. example: |- { "type": "", "context_id": "", "text": "", "voice_id": "", "speed": 123 } bindings: [] extensions: - id: x-parser-unique-object-id value: inputText - &ref_17 id: inputFlush payload: - name: Flush and generate description: >- Generate the context's buffered text as its next segment. The context stays open. type: object properties: - name: type type: string description: input.flush required: true - name: context_id type: string description: The context this message is for. examples: &ref_8 - turn-1 required: false headers: [] jsonPayloadSchema: type: object required: - type properties: type: type: string const: input.flush x-parser-schema-id: context_id: type: string pattern: ^[A-Za-z0-9_.:-]{1,128}$ default: default description: The context this message is for. examples: *ref_8 x-parser-schema-id: x-parser-schema-id: title: Flush and generate description: >- Generate the context's buffered text as its next segment. The context stays open. example: |- { "type": "", "context_id": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: inputFlush - &ref_18 id: contextClose payload: - name: Close context description: > No more text for this context. The rest of its buffer is generated, all audio is delivered, then context.done arrives with reason completed. type: object properties: - name: type type: string description: context.close required: true - name: context_id type: string description: The context this message is for. examples: &ref_9 - turn-1 required: false headers: [] jsonPayloadSchema: type: object required: - type properties: type: type: string const: context.close x-parser-schema-id: context_id: type: string pattern: ^[A-Za-z0-9_.:-]{1,128}$ default: default description: The context this message is for. examples: *ref_9 x-parser-schema-id: x-parser-schema-id: title: Close context description: > No more text for this context. The rest of its buffer is generated, all audio is delivered, then context.done arrives with reason completed. example: |- { "type": "", "context_id": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: contextClose - &ref_19 id: contextCancel payload: - name: Cancel context description: > End the context now: generation in progress is stopped, queued and unflushed text is discarded, and context.done arrives with reason cancelled. No audio for the context is sent after it. Ignored if the context has already ended. type: object properties: - name: type type: string description: context.cancel required: true - name: context_id type: string description: The context this message is for. examples: &ref_10 - turn-1 required: false headers: [] jsonPayloadSchema: type: object required: - type properties: type: type: string const: context.cancel x-parser-schema-id: context_id: type: string pattern: ^[A-Za-z0-9_.:-]{1,128}$ default: default description: The context this message is for. examples: *ref_10 x-parser-schema-id: x-parser-schema-id: title: Cancel context description: > End the context now: generation in progress is stopped, queued and unflushed text is discarded, and context.done arrives with reason cancelled. No audio for the context is sent after it. Ignored if the context has already ended. example: |- { "type": "", "context_id": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: contextCancel - &ref_20 id: sessionClose payload: - name: Close session description: >- Finish every open context as if each received context.close, then close the connection. type: object properties: - name: type type: string description: session.close required: true headers: [] jsonPayloadSchema: type: object required: - type properties: type: type: string const: session.close x-parser-schema-id: x-parser-schema-id: title: Close session description: >- Finish every open context as if each received context.close, then close the connection. example: |- { "type": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: sessionClose - &ref_21 id: ping payload: - name: Ping description: Keep the connection alive. Answered with pong. type: object properties: - name: type type: string description: ping required: true headers: [] jsonPayloadSchema: type: object required: - type properties: type: type: string const: ping x-parser-schema-id: x-parser-schema-id: title: Ping description: Keep the connection alive. Answered with pong. example: |- { "type": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: ping bindings: [] extensions: &ref_13 - id: x-parser-unique-object-id value: ttsMultiStream - &ref_15 id: receiveFromChariotMulti title: Server to client description: Messages Chariot sends back to you. type: send messages: - &ref_22 id: sessionReady payload: - name: Session ready description: The connection is authenticated and ready. type: object properties: - name: type type: string description: session.ready required: false - name: session_id type: string required: false - name: max_contexts type: integer examples: &ref_11 - 5 required: false - name: sample_rate type: integer examples: &ref_12 - 44100 required: false - name: encoding type: string description: pcm_s16le required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: session.ready x-parser-schema-id: session_id: type: string format: uuid x-parser-schema-id: max_contexts: type: integer examples: *ref_11 x-parser-schema-id: sample_rate: type: integer examples: *ref_12 x-parser-schema-id: encoding: type: string const: pcm_s16le x-parser-schema-id: x-parser-schema-id: title: Session ready description: The connection is authenticated and ready. example: |- { "type": "", "session_id": "", "max_contexts": 123, "sample_rate": 123, "encoding": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: sessionReady - &ref_23 id: audioStart payload: - name: Segment started description: A segment of a context is about to stream. type: object properties: - name: type type: string description: audio.start required: false - name: context_id type: string required: false - name: segment type: integer description: Zero-based segment number within the context. required: false - name: sample_rate type: integer required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: audio.start x-parser-schema-id: context_id: type: string x-parser-schema-id: segment: type: integer description: Zero-based segment number within the context. x-parser-schema-id: sample_rate: type: integer default: 44100 x-parser-schema-id: x-parser-schema-id: title: Segment started description: A segment of a context is about to stream. example: |- { "type": "", "context_id": "", "segment": 123, "sample_rate": 123 } bindings: [] extensions: - id: x-parser-unique-object-id value: audioStart - &ref_24 id: audioChunk payload: - name: Audio description: Audio for a segment. type: object properties: - name: type type: string description: audio.chunk required: false - name: context_id type: string required: false - name: segment type: integer required: false - name: audio type: string description: Base64-encoded raw PCM, signed 16-bit little-endian, mono. required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: audio.chunk x-parser-schema-id: context_id: type: string x-parser-schema-id: segment: type: integer x-parser-schema-id: audio: type: string contentEncoding: base64 description: Base64-encoded raw PCM, signed 16-bit little-endian, mono. x-parser-schema-id: x-parser-schema-id: title: Audio description: Audio for a segment. example: |- { "type": "", "context_id": "", "segment": 123, "audio": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: audioChunk - &ref_25 id: audioDone payload: - name: Segment finished description: >- The segment has finished streaming. Carries the credits charged for it. type: object properties: - name: type type: string description: audio.done required: false - name: context_id type: string required: false - name: segment type: integer required: false - name: credits_utilized type: integer description: >- Credits charged for this segment, at 1 credit per input character. required: false - name: error type: boolean required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: audio.done x-parser-schema-id: context_id: type: string x-parser-schema-id: segment: type: integer x-parser-schema-id: credits_utilized: type: integer description: >- Credits charged for this segment, at 1 credit per input character. x-parser-schema-id: error: type: boolean x-parser-schema-id: x-parser-schema-id: title: Segment finished description: >- The segment has finished streaming. Carries the credits charged for it. example: |- { "type": "", "context_id": "", "segment": 123, "credits_utilized": 123, "error": true } bindings: [] extensions: - id: x-parser-unique-object-id value: audioDone - &ref_26 id: contextDone payload: - name: Context ended description: The last message for a context, whatever ended it. type: object properties: - name: type type: string description: context.done required: false - name: context_id type: string required: false - name: reason type: string enumValues: - completed - cancelled - expired - error - session_closed required: false - name: segments type: integer description: Segments of this context that finished. required: false - name: credits_utilized type: integer description: > Total credits charged for this context. A segment that had started generating when the context was cancelled is included. required: false - name: code type: string description: Present when reason is error. required: false - name: message type: string description: Present when reason is error. required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: context.done x-parser-schema-id: context_id: type: string x-parser-schema-id: reason: type: string enum: - completed - cancelled - expired - error - session_closed x-parser-schema-id: segments: type: integer description: Segments of this context that finished. x-parser-schema-id: credits_utilized: type: integer description: > Total credits charged for this context. A segment that had started generating when the context was cancelled is included. x-parser-schema-id: code: type: string description: Present when reason is error. x-parser-schema-id: message: type: string description: Present when reason is error. x-parser-schema-id: x-parser-schema-id: title: Context ended description: The last message for a context, whatever ended it. example: |- { "type": "", "context_id": "", "reason": "", "segments": 123, "credits_utilized": 123, "code": "", "message": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: contextDone - &ref_27 id: errorMessage payload: - name: Error description: > A message was rejected. The connection stays open unless code is insufficient_credits (closes with 4402) or idle_timeout (closes with 1008). type: object properties: - name: type type: string description: error required: false - name: code type: string enumValues: - bad_request - unknown_context - context_limit - context_closing - session_closing - text_too_long - voice_unavailable - concurrency_limit - forbidden - upstream_error - insufficient_credits - idle_timeout required: false - name: message type: string required: false - name: context_id type: string description: The context the error concerns, if any. required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: error x-parser-schema-id: code: type: string enum: - bad_request - unknown_context - context_limit - context_closing - session_closing - text_too_long - voice_unavailable - concurrency_limit - forbidden - upstream_error - insufficient_credits - idle_timeout x-parser-schema-id: message: type: string x-parser-schema-id: context_id: type: string description: The context the error concerns, if any. x-parser-schema-id: x-parser-schema-id: title: Error description: > A message was rejected. The connection stays open unless code is insufficient_credits (closes with 4402) or idle_timeout (closes with 1008). example: |- { "type": "", "code": "", "message": "", "context_id": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: errorMessage - &ref_28 id: sessionDone payload: - name: Session complete description: >- Sent after session.close once every context has finished. The connection then closes with code 1000. type: object properties: - name: type type: string description: session.done required: false - name: contexts type: integer required: false - name: credits_utilized type: integer required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: session.done x-parser-schema-id: contexts: type: integer x-parser-schema-id: credits_utilized: type: integer x-parser-schema-id: x-parser-schema-id: title: Session complete description: >- Sent after session.close once every context has finished. The connection then closes with code 1000. example: |- { "type": "", "contexts": 123, "credits_utilized": 123 } bindings: [] extensions: - id: x-parser-unique-object-id value: sessionDone - &ref_29 id: pong payload: - name: Pong description: Answer to ping. type: object properties: - name: type type: string description: pong required: false headers: [] jsonPayloadSchema: type: object properties: type: type: string const: pong x-parser-schema-id: x-parser-schema-id: title: Pong description: Answer to ping. example: |- { "type": "" } bindings: [] extensions: - id: x-parser-unique-object-id value: pong bindings: [] extensions: *ref_13 sendOperations: - *ref_14 receiveOperations: - *ref_15 sendMessages: - *ref_16 - *ref_17 - *ref_18 - *ref_19 - *ref_20 - *ref_21 receiveMessages: - *ref_22 - *ref_23 - *ref_24 - *ref_25 - *ref_26 - *ref_27 - *ref_28 - *ref_29 extensions: - id: x-parser-unique-object-id value: ttsMultiStream securitySchemes: [] ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.