--- title: "Text To Speech with Timestamps" method: POST path: "/v1/text-to-speech/with-timestamps" tags: ["Text-to-Speech"] --- # Text To Speech with Timestamps `POST /v1/text-to-speech/with-timestamps` ## Request - Base URL: `https://api.typecast.ai` - URL: `https://api.typecast.ai/v1/text-to-speech/with-timestamps` - Auth: API key in header `X-API-KEY` Generate speech from text **and** return word/character-level timestamps aligned with the audio. Useful for subtitle sync, per-character highlight animation, and speech-region visualization. The request body matches the standard `/v1/text-to-speech` endpoint (voice_id, text, model, language, prompt, output). Instead of raw audio bytes, this endpoint returns a JSON object containing base64-encoded audio plus `words` and `characters` arrays. Use the optional `granularity` query parameter to return only word-level or only character-level timestamps and reduce payload size. > **Language note.** For languages that do not use whitespace between words — such as Japanese (`jpn`) and Chinese (`zho`) — word-level alignment collapses the entire sentence into a single "word". For those languages, always request `granularity=char` to receive usable per-character timestamps. See [Listing all voices](/docs/api-reference/voices/list-voices) for available voices. ## Query parameters - `granularity` 'word' | 'char' ## Request body - object: Text-to-speech request parameters - `text` string, required: Text to convert to speech. Minimum 1 character, maximum 2000 characters. Credits consumed based on text length. Supports multiple languages including English, Korean, Japanese, and Chinese. Special characters and punctuation are handled automatically. - `model` 'ssfm-v30' | 'ssfm-v21', required: TTS model version to use for speech synthesis. Different models offer varying capabilities and quality levels. Available models: - **ssfm-v30**: Latest model with improved prosody and additional emotion presets (recommended) - **ssfm-v21**: Stable production model with proven reliability and consistent quality - `output` object: Audio output settings including volume (0-200), pitch (-12 to +12 semitones), tempo (0.5x to 2.0x), and format (wav/mp3) for controlling the final audio characteristics Use `remove_silence_ms` (integer, 0–1000 ms) to shorten detected silence. - `volume` integer, nullable: Adjusts the relative volume of the output audio: 0 (completely silent), 50 (half volume), 100 (standard volume, default), 150 (50% louder than standard), 200 (maximum volume, twice as loud as standard). Since this only scales the existing volume, using `volume` can amplify the loudness differences between voices if they have different baseline levels. For consistent output across all clips, use `target_lufs` instead. - **Note:** This parameter cannot be used simultaneously with the `target_lufs` parameter. Required range: 0 <= x <= 200 - `audio_pitch` integer: Adjusts the pitch in semitones to affect perceived gender and age: -12 (one octave lower, deeper voice), -6 (half octave lower), 0 (original pitch, default), +6 (half octave higher), +12 (one octave higher, higher voice) - `audio_tempo` number: Controls speech speed: 0.5 (half speed, very slow and clear), 0.75 (slightly slower than normal), 1.0 (normal speaking speed, default), 1.5 (50% faster than normal), 2.0 (double speed, very fast speech) - `target_lufs` number, nullable: Sets the target absolute loudness (LUFS) for the output audio. This normalizes all generated voices to a consistent volume level, regardless of the original source's loudness. Values closer to 0 are louder, while values closer to -70 are quieter. - Required range: -70 <= x <= 0 - Recommended values: -14 (common streaming standard), -23 (broadcast standard) - **Note:** This parameter cannot be used simultaneously with the `volume` parameter. Use `target_lufs` for consistent absolute loudness across different clips, or use `volume` for traditional relative scaling. - `audio_format` 'wav' | 'mp3': Output audio format. **WAV format:** - Uncompressed PCM audio - 16-bit depth, mono channel, 44100 Hz sample rate - Higher quality, larger file size - Recommended for professional audio production **MP3 format:** - Compressed MPEG Layer III audio - 320 kbps bitrate, 44100 Hz sample rate - Smaller file size - Recommended for web streaming and distribution - `remove_silence_ms` integer, nullable: When enabled, shortens detected silences longer than the specified duration to that duration. The value is in milliseconds (ms). The value is the duration of silence to retain, not the amount to remove. **Accepted values:** - An integer from **0 to 1000**, with a recommended range of 0 to 200. - Omitted or `null`: silence removal is disabled. - `0`: removes detected silence longer than 0 ms. **This does not disable the feature.** - Booleans, strings, fractional values, and out-of-range values are invalid. **Example:** `100` shortens detected silences longer than 100 ms to 100 ms. Lower values include shorter silence segments for removal and leave less silence in each affected segment. - `prompt` union: Emotion and style settings for the generated speech, including emotion type (happy/sad/angry/normal) and intensity (0.0 to 2.0) to control the emotional expression - SmartPrompt: Emotion and style settings for the generated speech. - `next_text` string: Text that comes AFTER the `text` field in TTSRequest. Provides forward context for emotion inference. The model analyzes the flow: `previous_text` → `text` (synthesized) → `next_text` - Maximum 2000 characters - Helps the model anticipate emotional transitions - Leave empty if no following context is available - `emotion_type` 'smart': Discriminator field to identify the prompt type. Must be set to "smart" for context-aware emotion inference. - `previous_text` string: Text that comes BEFORE the `text` field in TTSRequest. Provides backward context for emotion inference. The model analyzes the flow: `previous_text` → `text` (synthesized) → `next_text` - Maximum 2000 characters - Helps the model understand emotional build-up and context - Leave empty if no preceding context is available - PresetPrompt: Emotion and style settings for the generated speech. - `emotion_type` 'preset': Discriminator field to identify the prompt type. Must be set to "preset" for preset-based emotion control. - `emotion_preset` 'normal' | 'sad' | 'happy' | 'angry' | 'whisper' | 'toneup' | 'tonedown': Available emotion presets for speech synthesis. Each emotion affects the tone, pace, and expressiveness of the generated speech. **ssfm-v21 Supported Emotions (4 types):** - normal: Neutral, balanced tone - happy: Bright, cheerful expression - sad: Melancholic, subdued tone - angry: Strong, intense delivery **ssfm-v30 Supported Emotions (7 types):** - normal: Neutral, balanced tone - happy: Bright, cheerful expression - sad: Melancholic, subdued tone - angry: Strong, intense delivery - whisper: Soft, quiet speech - toneup: Higher tonal emphasis - tonedown: Lower tonal emphasis Check available emotions for each voice in the `models` field of the `GET /v3/voices` API response. - `emotion_intensity` number: Controls the strength of emotional expression in the generated speech. - 0.0: Completely neutral, no emotional coloring - 0.5: Subtle emotional hints - 1.0: Standard emotional expression (default) - 1.5: Strong emotional emphasis - 2.0: Maximum intensity, highly expressive - Prompt: Emotion and style settings for the generated speech. - `emotion_preset` string: Emotion preset to apply. Supported emotions for ssfm-v21: normal, happy, sad, angry Check available emotions for each voice in the `models` field of the `GET /v3/voices` API response. - `emotion_intensity` unknown - `language` string: Language code following ISO 639-3 standard. Case-insensitive (both "ENG" and "eng" are accepted). If not provided, will be auto-detected based on text content.
ssfm-v30 Supported Languages (37) | Code | Language | Code | Language | Code | Language | | ---- | --------- | ---- | ---------- | ---- | ---------- | | ARA | Arabic | IND | Indonesian | POR | Portuguese | | BEN | Bengali | ITA | Italian | RON | Romanian | | BUL | Bulgarian | JPN | Japanese | RUS | Russian | | CES | Czech | KOR | Korean | SLK | Slovak | | DAN | Danish | MSA | Malay | SPA | Spanish | | DEU | German | NAN | Min Nan | SWE | Swedish | | ELL | Greek | NLD | Dutch | TAM | Tamil | | ENG | English | NOR | Norwegian | TGL | Tagalog | | FIN | Finnish | PAN | Punjabi | THA | Thai | | FRA | French | POL | Polish | TUR | Turkish | | HIN | Hindi | UKR | Ukrainian | VIE | Vietnamese | | HRV | Croatian | YUE | Cantonese | ZHO | Chinese | | HUN | Hungarian | | | | |
ssfm-v21 Supported Languages (27) | Code | Language | Code | Language | Code | Language | | ---- | --------- | ---- | ---------- | ---- | --------- | | ARA | Arabic | IND | Indonesian | RON | Romanian | | BUL | Bulgarian | ITA | Italian | RUS | Russian | | CES | Czech | JPN | Japanese | SLK | Slovak | | DAN | Danish | KOR | Korean | SPA | Spanish | | DEU | German | MSA | Malay | SWE | Swedish | | ELL | Greek | NLD | Dutch | TAM | Tamil | | ENG | English | POL | Polish | TGL | Tagalog | | FIN | Finnish | POR | Portuguese | UKR | Ukrainian | | FRA | French | HRV | Croatian | ZHO | Chinese |
> **Timestamp endpoint note.** For languages without inter-word whitespace — Japanese (`jpn`) and Chinese (`zho`) — word-level alignment collapses the whole sentence into a single segment. Always pair these languages with `granularity=char` to receive usable per-character timestamps. - `voice_id` string, required: Voice identifier. Two prefixes are supported: * `tc_` — Built-in Typecast voices (e.g., `tc_60e5426de8b95f1d3000d7b5`). See [Listing all voices](/docs/api-reference/voices/list-voices) for available IDs. * `uc_` — Custom voices created via [Instant cloning](/docs/api-reference/voices/instant-cloning) (e.g., `uc_64a1b2c3d4e5f6a7b8c9d0e1`). Only the owner of a cloned voice can use it. Case-sensitive: must use lowercase prefix. ## Response `200` Success - Returns base64 audio and timestamps - TTSWithTimestampsResponse: Response payload for POST /v1/text-to-speech/with-timestamps — base64-encoded audio plus per-word and per-character timestamps aligned with the generated speech. - `audio` string, required: Base64-encoded audio bytes. Decode and write to a file using the `audio_format` extension. - `words` AlignmentSegmentWord[], nullable, required: Word-level timestamps (with attached punctuation). `null` when the request uses `granularity=char`. - `end` number, required: End time of this segment, in seconds from the beginning of the audio. - `text` string, required: The text fragment from the original transcript (includes any attached punctuation). - `start` number, required: Start time of this segment, in seconds from the beginning of the audio. - `characters` AlignmentSegmentCharacter[], nullable, required: Character-level timestamps (including punctuation and whitespace). `null` when the request uses `granularity=word`. - `end` number, required: End time of this segment, in seconds from the beginning of the audio. - `text` string, required: The text fragment from the original transcript (includes any attached punctuation and whitespace). - `start` number, required: Start time of this segment, in seconds from the beginning of the audio. - `audio_format` 'wav' | 'mp3', required: Audio encoding format of the bytes in `audio` — either `wav` or `mp3`, mirroring the request's `output.audio_format`. - `audio_duration` number, required: Length of the generated audio in seconds. ## Other responses - `400`: Bad Request - Invalid parameters - `401`: Unauthorized - Authentication failed - `402`: Payment Required - Insufficient credits - `404`: Not Found - Voice model not available - `422`: Validation Error - The request is invalid or the input text cannot be synthesized - `429`: Too Many Requests - Rate limit exceeded - `500`: Internal Server Error - TTS generation or timestamp alignment failed --- [API](https://skmtc.dev/typecast/apis/typecast-api.md) · [All operations](https://skmtc.dev/typecast/apis/typecast-api/llms.txt) · [OpenAPI document](https://skmtc.dev/typecast/apis/typecast-api/revisions/41605154de82?raw)