> ## 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. # Kugel 1 > Stream high-quality speech synthesis from KugelAudio Kugel 1 over WebSocket with expressiveness controls and SLNG's unified TTS protocol. ## AsyncAPI ````yaml api-reference/asyncapi/tts-kugelaudio.asyncapi.yaml /v1/tts/kugelaudio/kugel:1 id: /v1/tts/kugelaudio/kugel:1 title: Kugel 1 description: >- Stream high-quality speech synthesis from KugelAudio Kugel 1 over WebSocket with expressiveness controls and SLNG's unified TTS protocol. servers: - id: production protocol: wss host: api.slng.ai bindings: [] variables: [] - id: staging protocol: wss host: stageapi.slng.ai bindings: [] variables: [] address: /v1/tts/kugelaudio/kugel:1 parameters: [] bindings: - protocol: ws version: latest value: method: GET headers: type: object properties: X-World-Part-Override: type: string description: >- Target world part override. Auto-selected if not provided. Available world parts: `eu`. enum: - eu x-parser-schema-id: x-parser-schema-id: WebsocketHeadersKugelAudio schemaProperties: - name: method type: string description: GET required: false - name: headers type: object required: false properties: - name: X-World-Part-Override type: string description: >- Target world part override. Auto-selected if not provided. Available world parts: `eu`. enumValues: - eu required: false operations: - &ref_1 id: ttsKugelAudioKugel1ReceiveRequest title: Tts kugel audio kugel1 receive request description: Send TTS request to Kugel 1 type: receive messages: - &ref_13 id: KugelTtsRequest contentType: application/json payload: - name: Request description: Send text with synthesis parameters to KugelAudio TTS. type: object properties: - name: type type: string description: request required: true - name: text type: string description: Text to convert to speech required: true - name: model_id type: string description: Model selection enumValues: - kugel-2-turbo - kugel-1-turbo - kugel-1 - kugel-2 required: false - name: voice_id type: integer description: Voice identifier required: false - name: cfg_scale type: number description: >- Classifier-free guidance scale. Higher values = more expressive required: false - name: normalize type: boolean description: Enable text normalization required: false - name: language type: string description: ISO 639-1 language code for text normalization enumValues: - de - en - fr - es - it - pt - nl - pl - sv - da - 'no' - fi - cs - hu - ro - el - uk - bg - tr - vi - ar - hi - zh - ja - ko required: false - name: speed type: number description: Playback speed multiplier (pitch-preserving WSOLA) required: false - name: word_timestamps type: boolean description: Enable word-level timestamp alignment required: false - name: speaker_prefix type: boolean description: Prepend internal speaker prefix for better voice consistency required: false - name: max_new_tokens type: integer description: Maximum tokens to generate required: false - name: sample_rate type: integer description: Output sample rate in Hz enumValues: - 8000 - 16000 - 22050 - 24000 required: false headers: [] jsonPayloadSchema: type: object description: >- Send text to convert to speech with all synthesis parameters in a single request message. Matches KugelAudio's native WebSocket API. required: - type - text properties: type: type: string const: request default: request x-parser-schema-id: text: type: string maxLength: 10000 description: Text to convert to speech x-parser-schema-id: model_id: type: string description: Model selection enum: - kugel-2-turbo - kugel-1-turbo - kugel-1 - kugel-2 default: kugel-1-turbo x-parser-schema-id: voice_id: type: integer description: Voice identifier x-parser-schema-id: cfg_scale: type: number minimum: 0 maximum: 10 default: 2 description: Classifier-free guidance scale. Higher values = more expressive x-parser-schema-id: normalize: type: boolean default: true description: Enable text normalization x-parser-schema-id: language: type: string description: ISO 639-1 language code for text normalization enum: - de - en - fr - es - it - pt - nl - pl - sv - da - 'no' - fi - cs - hu - ro - el - uk - bg - tr - vi - ar - hi - zh - ja - ko x-parser-schema-id: speed: type: number minimum: 0.8 maximum: 1.2 default: 1 description: Playback speed multiplier (pitch-preserving WSOLA) x-parser-schema-id: word_timestamps: type: boolean default: false description: Enable word-level timestamp alignment x-parser-schema-id: speaker_prefix: type: boolean default: true description: Prepend internal speaker prefix for better voice consistency x-parser-schema-id: max_new_tokens: type: integer default: 2048 minimum: 1 maximum: 8192 description: Maximum tokens to generate x-parser-schema-id: sample_rate: type: integer description: Output sample rate in Hz enum: - 8000 - 16000 - 22050 - 24000 default: 24000 x-parser-schema-id: x-parser-schema-id: title: Request description: Send text with synthesis parameters to KugelAudio TTS. example: |- { "type": "request", "text": "Hello from sunny Barcelona", "model_id": "kugel-1-turbo", "voice_id": 268, "cfg_scale": 2, "sample_rate": 24000 } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsRequest bindings: [] extensions: &ref_0 - id: x-parser-unique-object-id value: /v1/tts/kugelaudio/kugel:1 - &ref_2 id: ttsKugelAudioKugel1ReceiveConfig title: Tts kugel audio kugel1 receive config description: Configure streaming session on Kugel 1 type: receive messages: - &ref_14 id: KugelTtsConfigMessage contentType: application/json payload: - name: Config description: >- Configure a KugelAudio streaming TTS session before sending text chunks. type: object properties: - name: type type: string description: config required: true - name: voice_id type: integer description: Voice identifier required: false - name: model_id type: string description: Model selection enumValues: - kugel-2-turbo - kugel-1-turbo - kugel-1 - kugel-2 required: false - name: cfg_scale type: number description: >- Classifier-free guidance scale. Higher values = more expressive required: false - name: sample_rate type: integer description: Output sample rate in Hz enumValues: - 8000 - 16000 - 22050 - 24000 required: false - name: normalize type: boolean description: Enable text normalization required: false - name: language type: string description: ISO 639-1 language code for text normalization enumValues: - de - en - fr - es - it - pt - nl - pl - sv - da - 'no' - fi - cs - hu - ro - el - uk - bg - tr - vi - ar - hi - zh - ja - ko required: false - name: word_timestamps type: boolean description: Enable word-level timestamp alignment required: false - name: flush_timeout_ms type: integer description: Milliseconds before auto-flushing buffered text required: false - name: speed type: number description: Playback speed multiplier (pitch-preserving WSOLA) required: false headers: [] jsonPayloadSchema: type: object description: >- Initial configuration message for stream input mode. Sent once at the start of a session to set voice, model, and synthesis parameters before streaming text. required: - type properties: type: type: string const: config default: config x-parser-schema-id: voice_id: type: integer description: Voice identifier x-parser-schema-id: model_id: type: string description: Model selection enum: - kugel-2-turbo - kugel-1-turbo - kugel-1 - kugel-2 default: kugel-1-turbo x-parser-schema-id: cfg_scale: type: number minimum: 0 maximum: 10 default: 2 description: Classifier-free guidance scale. Higher values = more expressive x-parser-schema-id: sample_rate: type: integer description: Output sample rate in Hz enum: - 8000 - 16000 - 22050 - 24000 default: 24000 x-parser-schema-id: normalize: type: boolean default: true description: Enable text normalization x-parser-schema-id: language: type: string description: ISO 639-1 language code for text normalization enum: - de - en - fr - es - it - pt - nl - pl - sv - da - 'no' - fi - cs - hu - ro - el - uk - bg - tr - vi - ar - hi - zh - ja - ko x-parser-schema-id: word_timestamps: type: boolean default: false description: Enable word-level timestamp alignment x-parser-schema-id: flush_timeout_ms: type: integer default: 500 description: Milliseconds before auto-flushing buffered text x-parser-schema-id: speed: type: number minimum: 0.8 maximum: 1.2 default: 1 description: Playback speed multiplier (pitch-preserving WSOLA) x-parser-schema-id: x-parser-schema-id: title: Config description: >- Configure a KugelAudio streaming TTS session before sending text chunks. example: |- { "type": "config", "voice_id": 268, "model_id": "kugel-1-turbo", "cfg_scale": 2, "sample_rate": 24000, "flush_timeout_ms": 500 } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsConfigMessage bindings: [] extensions: *ref_0 - &ref_3 id: ttsKugelAudioKugel1ReceiveText title: Tts kugel audio kugel1 receive text description: Send text chunk to Kugel 1 type: receive messages: - &ref_15 id: KugelTtsTextMessage contentType: application/json payload: - name: Text Chunk description: Send an incremental text chunk for streaming TTS generation. type: object properties: - name: type type: string description: text required: true - name: text type: string description: Text chunk to buffer for synthesis required: true headers: [] jsonPayloadSchema: type: object description: >- Incremental text chunk for stream input mode. Send as tokens arrive from an LLM or other source. Text is buffered until flushed or flush_timeout_ms expires. required: - type - text properties: type: type: string const: text default: text x-parser-schema-id: text: type: string description: Text chunk to buffer for synthesis x-parser-schema-id: x-parser-schema-id: title: Text Chunk description: Send an incremental text chunk for streaming TTS generation. example: |- { "type": "text", "text": "Hello from sunny " } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsTextMessage bindings: [] extensions: *ref_0 - &ref_4 id: ttsKugelAudioKugel1ReceiveFlush title: Tts kugel audio kugel1 receive flush description: Flush buffered text on Kugel 1 type: receive messages: - &ref_16 id: KugelTtsFlushMessage contentType: application/json payload: - name: Flush description: Force immediate generation of all buffered text. type: object properties: - name: type type: string description: flush required: true - name: flush type: boolean description: Flush signal required: true headers: [] jsonPayloadSchema: type: object description: >- Forces immediate generation of all buffered text without waiting for flush_timeout_ms. required: - type - flush properties: type: type: string const: flush default: flush x-parser-schema-id: flush: type: boolean const: true description: Flush signal x-parser-schema-id: x-parser-schema-id: title: Flush description: Force immediate generation of all buffered text. example: |- { "type": "flush", "flush": true } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsFlushMessage bindings: [] extensions: *ref_0 - &ref_5 id: ttsKugelAudioKugel1ReceiveClose title: Tts kugel audio kugel1 receive close description: Close streaming session on Kugel 1 type: receive messages: - &ref_17 id: KugelTtsCloseMessage contentType: application/json payload: - name: Close description: End the streaming session and flush remaining buffered text. type: object properties: - name: type type: string description: close required: true - name: close type: boolean description: Close signal required: true headers: [] jsonPayloadSchema: type: object description: >- Terminates the streaming session after flushing any remaining buffered content. required: - type - close properties: type: type: string const: close default: close x-parser-schema-id: close: type: boolean const: true description: Close signal x-parser-schema-id: x-parser-schema-id: title: Close description: End the streaming session and flush remaining buffered text. example: |- { "type": "close", "close": true } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsCloseMessage bindings: [] extensions: *ref_0 - &ref_6 id: ttsKugelAudioKugel1SendAudioChunk title: Tts kugel audio kugel1 send audio chunk description: Receive audio chunk from Kugel 1 type: send messages: - &ref_18 id: KugelTtsAudioChunk contentType: application/json payload: - name: Audio Chunk description: >- Chunk of base64-encoded PCM audio from KugelAudio with encoding metadata. type: object properties: - name: type type: string description: audio_chunk required: true - name: audio type: string description: Base64-encoded PCM16 audio data required: true - name: enc type: string description: Audio encoding format required: true - name: idx type: integer description: Chunk index (0-based) required: true - name: sr type: integer description: Sample rate in Hz required: true - name: samples type: integer description: Number of samples in this chunk required: true - name: chunk_id type: integer description: >- Text chunk identifier (stream input mode). Links audio to the text chunk that generated it. required: false headers: [] jsonPayloadSchema: type: object description: >- Chunk of base64-encoded PCM audio from KugelAudio with encoding metadata. required: - type - audio - enc - idx - sr - samples properties: type: type: string const: audio_chunk default: audio_chunk x-parser-schema-id: audio: type: string format: byte description: Base64-encoded PCM16 audio data x-parser-schema-id: enc: type: string const: pcm_s16le description: Audio encoding format x-parser-schema-id: idx: type: integer description: Chunk index (0-based) x-parser-schema-id: sr: type: integer description: Sample rate in Hz x-parser-schema-id: samples: type: integer description: Number of samples in this chunk x-parser-schema-id: chunk_id: type: integer description: >- Text chunk identifier (stream input mode). Links audio to the text chunk that generated it. x-parser-schema-id: x-parser-schema-id: title: Audio Chunk description: >- Chunk of base64-encoded PCM audio from KugelAudio with encoding metadata. example: |- { "type": "audio_chunk", "audio": "UklGRiQAAABXQVZFZm10IBAAAAABAAEA...", "enc": "pcm_s16le", "idx": 0, "sr": 24000, "samples": 4800, "chunk_id": 0 } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsAudioChunk bindings: [] extensions: *ref_0 - &ref_7 id: ttsKugelAudioKugel1SendWordTimestamps title: Tts kugel audio kugel1 send word timestamps description: Receive word timestamps from Kugel 1 type: send messages: - &ref_19 id: KugelTtsWordTimestamps contentType: application/json payload: - name: Word Timestamps description: Word-level timestamp alignment data from KugelAudio. type: object properties: - name: type type: string description: word_timestamps required: true - name: word_timestamps type: array description: Array of word-level alignment entries required: true properties: - name: word type: string description: The spoken word required: true - name: start_ms type: number description: Word start time in milliseconds required: true - name: end_ms type: number description: Word end time in milliseconds required: true - name: char_start type: integer description: Start character position in input text required: true - name: char_end type: integer description: End character position in input text required: true - name: score type: number description: Alignment confidence score required: true - name: chunk_id type: integer description: >- Text chunk identifier (stream input mode). Links timestamps to the text chunk that generated them. required: false headers: [] jsonPayloadSchema: type: object description: >- Word-level timestamp alignment data sent when word_timestamps is enabled. required: - type - word_timestamps properties: type: type: string const: word_timestamps default: word_timestamps x-parser-schema-id: word_timestamps: type: array description: Array of word-level alignment entries items: type: object required: - word - start_ms - end_ms - char_start - char_end - score properties: word: type: string description: The spoken word x-parser-schema-id: start_ms: type: number description: Word start time in milliseconds x-parser-schema-id: end_ms: type: number description: Word end time in milliseconds x-parser-schema-id: char_start: type: integer description: Start character position in input text x-parser-schema-id: char_end: type: integer description: End character position in input text x-parser-schema-id: score: type: number description: Alignment confidence score x-parser-schema-id: x-parser-schema-id: x-parser-schema-id: chunk_id: type: integer description: >- Text chunk identifier (stream input mode). Links timestamps to the text chunk that generated them. x-parser-schema-id: x-parser-schema-id: title: Word Timestamps description: Word-level timestamp alignment data from KugelAudio. example: |- { "type": "word_timestamps", "word_timestamps": [ { "word": "Hello", "start_ms": 0, "end_ms": 320, "char_start": 0, "char_end": 5, "score": 0.98 }, { "word": "world", "start_ms": 340, "end_ms": 620, "char_start": 6, "char_end": 11, "score": 0.95 } ], "chunk_id": 0 } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsWordTimestamps bindings: [] extensions: *ref_0 - &ref_8 id: ttsKugelAudioKugel1SendFinal title: Tts kugel audio kugel1 send final description: Receive final response from Kugel 1 type: send messages: - &ref_20 id: KugelTtsFinalResponse contentType: application/json payload: - name: Final Response description: >- Signals generation is complete with KugelAudio performance statistics. type: object properties: - name: type type: string description: final required: true - name: final type: boolean description: Indicates generation is complete required: true - name: chunks type: integer description: Number of audio chunks generated required: false - name: total_samples type: integer description: Total audio samples generated required: false - name: dur_ms type: number description: Total audio duration in milliseconds required: false - name: gen_ms type: number description: Total generation time in milliseconds required: false - name: rtf type: number description: Real-time factor (gen_ms / dur_ms) required: false headers: [] jsonPayloadSchema: type: object description: >- Signals generation is complete with performance statistics from KugelAudio. required: - type - final properties: type: type: string const: final default: final x-parser-schema-id: final: type: boolean const: true description: Indicates generation is complete x-parser-schema-id: chunks: type: integer description: Number of audio chunks generated x-parser-schema-id: total_samples: type: integer description: Total audio samples generated x-parser-schema-id: dur_ms: type: number description: Total audio duration in milliseconds x-parser-schema-id: gen_ms: type: number description: Total generation time in milliseconds x-parser-schema-id: rtf: type: number description: Real-time factor (gen_ms / dur_ms) x-parser-schema-id: x-parser-schema-id: title: Final Response description: Signals generation is complete with KugelAudio performance statistics. example: |- { "type": "final", "final": true, "chunks": 10, "total_samples": 48000, "dur_ms": 2000, "gen_ms": 150, "rtf": 0.075 } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsFinalResponse bindings: [] extensions: *ref_0 - &ref_9 id: ttsKugelAudioKugel1SendGenerationStarted title: Tts kugel audio kugel1 send generation started description: Receive generation started from Kugel 1 type: send messages: - &ref_21 id: KugelTtsGenerationStarted contentType: application/json payload: - name: Generation Started description: Indicates buffered text has reached the generation threshold. type: object properties: - name: type type: string description: generation_started required: true - name: generation_started type: boolean description: Generation started signal required: true - name: chunk_id type: integer description: Text chunk identifier (0-based) required: true - name: text type: string description: The text being generated required: true headers: [] jsonPayloadSchema: type: object description: >- Sent when buffered text reaches the generation threshold and audio synthesis begins for a text chunk. required: - type - generation_started - chunk_id - text properties: type: type: string const: generation_started default: generation_started x-parser-schema-id: generation_started: type: boolean const: true description: Generation started signal x-parser-schema-id: chunk_id: type: integer description: Text chunk identifier (0-based) x-parser-schema-id: text: type: string description: The text being generated x-parser-schema-id: x-parser-schema-id: title: Generation Started description: Indicates buffered text has reached the generation threshold. example: |- { "type": "generation_started", "generation_started": true, "chunk_id": 0, "text": "Hello from sunny Barcelona" } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsGenerationStarted bindings: [] extensions: *ref_0 - &ref_10 id: ttsKugelAudioKugel1SendChunkComplete title: Tts kugel audio kugel1 send chunk complete description: Receive chunk complete from Kugel 1 type: send messages: - &ref_22 id: KugelTtsChunkComplete contentType: application/json payload: - name: Chunk Complete description: Signals that audio generation for a text chunk is complete. type: object properties: - name: type type: string description: chunk_complete required: true - name: chunk_complete type: boolean description: Chunk complete signal required: true - name: chunk_id type: integer description: Text chunk identifier required: true - name: audio_seconds type: number description: Duration of audio generated for this chunk required: false - name: gen_ms type: number description: Generation time in milliseconds for this chunk required: false - name: is_final type: boolean description: Whether this is the last chunk in the session required: false headers: [] jsonPayloadSchema: type: object description: >- Sent when audio generation for an individual text chunk finishes. Includes performance metrics for the chunk. required: - type - chunk_complete - chunk_id properties: type: type: string const: chunk_complete default: chunk_complete x-parser-schema-id: chunk_complete: type: boolean const: true description: Chunk complete signal x-parser-schema-id: chunk_id: type: integer description: Text chunk identifier x-parser-schema-id: audio_seconds: type: number description: Duration of audio generated for this chunk x-parser-schema-id: gen_ms: type: number description: Generation time in milliseconds for this chunk x-parser-schema-id: is_final: type: boolean description: Whether this is the last chunk in the session x-parser-schema-id: x-parser-schema-id: title: Chunk Complete description: Signals that audio generation for a text chunk is complete. example: |- { "type": "chunk_complete", "chunk_complete": true, "chunk_id": 0, "audio_seconds": 1.2, "gen_ms": 150, "is_final": false } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsChunkComplete bindings: [] extensions: *ref_0 - &ref_11 id: ttsKugelAudioKugel1SendSessionClosed title: Tts kugel audio kugel1 send session closed description: Receive session closed from Kugel 1 type: send messages: - &ref_23 id: KugelTtsSessionClosed contentType: application/json payload: - name: Session Closed description: Signals the streaming session has ended with aggregate statistics. type: object properties: - name: type type: string description: session_closed required: true - name: session_closed type: boolean description: Session closed signal required: true - name: total_audio_seconds type: number description: Cumulative audio duration across all chunks required: false - name: total_text_chunks type: integer description: Total number of text chunks processed required: false - name: total_audio_chunks type: integer description: Total number of audio chunks generated required: false headers: [] jsonPayloadSchema: type: object description: >- Sent after all text chunks have been processed and the session is terminated. Contains aggregate performance statistics. required: - type - session_closed properties: type: type: string const: session_closed default: session_closed x-parser-schema-id: session_closed: type: boolean const: true description: Session closed signal x-parser-schema-id: total_audio_seconds: type: number description: Cumulative audio duration across all chunks x-parser-schema-id: total_text_chunks: type: integer description: Total number of text chunks processed x-parser-schema-id: total_audio_chunks: type: integer description: Total number of audio chunks generated x-parser-schema-id: x-parser-schema-id: title: Session Closed description: Signals the streaming session has ended with aggregate statistics. example: |- { "type": "session_closed", "session_closed": true, "total_audio_seconds": 5.4, "total_text_chunks": 3, "total_audio_chunks": 15 } bindings: [] extensions: - id: x-parser-unique-object-id value: KugelTtsSessionClosed bindings: [] extensions: *ref_0 - &ref_12 id: ttsKugelAudioKugel1SendError title: Tts kugel audio kugel1 send error description: Receive error from Kugel 1 type: send messages: - &ref_24 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: code type: string description: Error code. enumValues: - auth_error - config_error - rate_limit - provider_error required: true - name: message type: string description: Human-readable error description required: true headers: [] jsonPayloadSchema: type: object description: Indicates an error occurred during synthesis. required: - type - code - message properties: type: type: string const: error x-parser-schema-id: code: type: string description: Error code. enum: - auth_error - config_error - rate_limit - provider_error x-parser-schema-id: message: type: string description: Human-readable error description x-parser-schema-id: x-parser-schema-id: title: Error Response description: Indicates an error occurred during synthesis. example: |- { "type": "error", "code": "provider_error", "message": "Provider returned an unexpected error" } bindings: [] extensions: - id: x-parser-unique-object-id value: TtsErrorResponse bindings: [] extensions: *ref_0 sendOperations: - *ref_1 - *ref_2 - *ref_3 - *ref_4 - *ref_5 receiveOperations: - *ref_6 - *ref_7 - *ref_8 - *ref_9 - *ref_10 - *ref_11 - *ref_12 sendMessages: - *ref_13 - *ref_14 - *ref_15 - *ref_16 - *ref_17 receiveMessages: - *ref_18 - *ref_19 - *ref_20 - *ref_21 - *ref_22 - *ref_23 - *ref_24 extensions: - id: x-parser-unique-object-id value: /v1/tts/kugelaudio/kugel:1 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: [] ````