ElevenLabs Speech-to-Text
AI-generatedOverview
This node converts audio files into text using the ElevenLabs Speech-to-Text API. It supports transcription from either a binary audio file or an audio URL. The node is useful for automating transcription workflows, such as converting meeting recordings, podcasts, interviews, or any audio content into searchable and editable text.
Practical examples include:
- Transcribing customer support calls to analyze conversations.
- Generating subtitles or captions for videos.
- Creating searchable archives of audio content.
- Identifying speakers in multi-person recordings (diarization).
Properties
| Name | Meaning |
|---|---|
| Fonte do Áudio | Source of the audio file for transcription. Options: "Arquivo Binário" (binary file) or "URL" (audio URL). |
| Campo Binário | Name of the binary property containing the audio file. Required if the source is a binary file. |
| URL do Áudio | URL of the audio file to transcribe. Required if the source is a URL. |
| Modelo | ID of the transcription model to use. Options: "Scribe v1" or "Scribe v1 Experimental". |
| Opções Avançadas | Collection of advanced options: |
| - Código de Idioma | ISO-639-1 or ISO-639-3 language code of the audio. If empty, language detection is automatic. |
| - Diarização | Boolean to enable speaker diarization (identifying different speakers). |
| - Timestamps | Granularity of timestamps in the transcript. Options: "Nenhum" (none), "Palavra" (word), "Caractere" (character). |
| - Marcar Eventos de Áudio | Boolean to tag audio events like (laughter), (footsteps), etc. |
| - Número Máximo de Falantes | Maximum number of speakers in the audio (1-32). |
| - Formato de Arquivo | Format of the input audio. Options: "Outro" (other), "PCM 16-bit 16kHz". |
| - Formatos Adicionais | Additional export formats for the transcription. Multiple values allowed, each with: |
| -- Tipo | Format type: DOCX, HTML, PDF, JSON Segmentado, SRT, TXT. |
| -- Incluir Falantes | Include speaker identification in the additional format. |
| -- Incluir Timestamps | Include timestamps in the additional format. |
| -- Máximo de Caracteres por Linha | Max characters per line (0 for no limit). |
| -- Máximo de Caracteres por Segmento | Max characters per segment (0 for no limit). |
| -- Duração Máxima do Segmento (s) | Max duration of each segment in seconds (0 for no limit). |
| -- Segmentar em Silêncios Maiores que (s) | Create new segment when silence longer than specified seconds is detected (0 to disable). |
Output
The node outputs JSON data containing the transcription results returned by the ElevenLabs API. This includes the transcribed text and may contain metadata such as timestamps, speaker labels (if diarization is enabled), and audio event tags depending on the selected options.
If additional export formats are requested, the output will also include these transcriptions in the specified formats.
The node does not output binary data directly; it processes audio input but returns transcription results as JSON.
Dependencies
- Requires an active ElevenLabs API key credential configured in n8n.
- Uses the ElevenLabs Speech-to-Text API endpoint via HTTP POST requests.
- No other external dependencies are required.
Troubleshooting
- No binary data found error: Occurs if the specified binary property does not exist or is empty when using binary file input. Ensure the correct binary property name is set and that the input contains valid audio data.
- Invalid URL or inaccessible audio URL: When using URL as audio source, ensure the URL is publicly accessible and points to a valid audio file.
- API errors: Errors from the ElevenLabs API (e.g., invalid API key, unsupported audio format) will be surfaced. Verify API credentials and input parameters.
- Unsupported audio format: If the audio format is not supported or incorrectly specified, transcription may fail. Use the "Formato de Arquivo" option to specify the correct format if needed.
- Exceeding limits: Large audio files or too many speakers may cause issues. Adjust parameters accordingly.