saidwire

n8n community node for Saidwire speech-to-text: transcribe English audio to text, timestamps or SRT/VTT subtitles.

Package Information

Released: this month
Downloads: 0 weekly / 0 monthly
Latest Version: 0.1.0
Author: Saidwire

Documentation

n8n-nodes-saidwire

This is an n8n community node. It lets you use Saidwire speech-to-text in your n8n workflows.

Saidwire transcribes English audio to text, segment and word timestamps, or SRT/WebVTT subtitles.
It costs $0.30 per audio hour, billed per second, and new keys get 120 free hours with no card
(saidwire.com/dev).

n8n is a fair-code licensed workflow automation platform.

Installation ·
Operations ·
Credentials ·
Compatibility ·
Usage ·
Resources ·
Version history

Installation

Follow the installation guide in the n8n community nodes documentation:

  1. Go to Settings > Community Nodes.
  2. Select Install.
  3. Enter n8n-nodes-saidwire and confirm.

On a self-hosted instance without the UI option, run npm install n8n-nodes-saidwire in ~/.n8n/nodes
and restart n8n.

Operations

Audio > Transcribe ("Transcribe audio")

Parameter Description
Input Binary Field Name of the binary field that holds the audio (default data)
Response Format JSON (Text Only), Verbose JSON (Segments and Timestamps), Plain Text, SRT Subtitles, WebVTT Subtitles
Options > Language en (default), a regional code such as en-US, or empty. English only.
Options > Timestamp Granularities Segment and/or Word (Verbose JSON only)
Options > Put Output in File Also return Plain Text, SRT or WebVTT output as a binary file
Options > Output Binary Field Binary field for that file (default data)

Output:

  • JSON: { "text": "...", "usage": { "type": "duration", "seconds": 12.0 } }
  • Verbose JSON: { "task", "language", "duration", "text", "segments": [{ "id", "start", "end", "text", ... }], "words": [{ "word", "start", "end" }], "usage" }
  • Plain Text: { "text": "..." }
  • SRT / WebVTT: { "subtitles": "...", "format": "srt" }, plus a .srt / .vtt binary file when Put Output in File is on

Audio files can be WAV, MP3, FLAC, Ogg/Opus, WebM or M4A/MP4, up to 100 MB and 60 minutes. For M4A/MP4, the file
index must be at the start of the file (for ffmpeg output, add -movflags +faststart).

Credentials

  1. Get an API key (sk_live_...) at saidwire.com/dev.
  2. In n8n, create a Saidwire API credential and paste the key into API Key.
    Leave Base URL as https://api.saidwire.com/v1.
  3. Save. n8n checks the key by listing the account's models (GET /v1/models).

The key is sent as Authorization: Bearer <key>.

Compatibility

  • Built with @n8n/node-cli 0.54 and n8n-workflow 2.x. It uses the standard httpRequestWithAuthentication helper and has no runtime dependencies.
  • Needs a recent n8n 1.x or 2.x with community nodes enabled.

Usage

A typical workflow:

  1. A node that outputs audio as binary data: Read/Write Files from Disk, an HTTP Request that downloads a
    file (Response Format: File), a Form Trigger with a file field, or a Google Drive / Dropbox download.
  2. Saidwire > Transcribe audio, with Input Binary Field set to that node's binary field (usually data).
  3. Use {{ $json.text }} in the next node, for example in an AI Agent, an email or a database row.

For subtitles, choose SRT Subtitles, turn on Put Output in File, and pass the binary on to a
Write Files to Disk or upload node.

Each input item is transcribed separately. Trial keys allow 4 files at once, so for large batches add a
Loop Over Items node with a small batch size. With On Error > Continue, failed items come back as
{ "error": "..." } and the rest of the batch keeps going.

Errors:

Message Fix
401 Incorrect API key provided Check the key in the credential
400 ... unsupported_language English only: set Language to en or leave it empty
400 ... file The binary field is empty or not audio. Check Input Binary Field
429 Too many files at once, or the free hours are used up (the message says which)

Resources

Saidwire's transcription endpoint is compatible with OpenAI's transcription API. OpenAI is a trademark of OpenAI.
n8n is a trademark of n8n GmbH. Saidwire is not affiliated with OpenAI or n8n.

Development

npm install
npm run lint        # n8n community node rules (strict, Cloud-eligible config)
npm test            # builds, then runs unit tests against a mocked n8n context
SAIDWIRE_LIVE=1 SAIDWIRE_API_KEY=sk_live_... SAIDWIRE_TEST_CLIP=/path/to/english.wav npm test   # adds live API calls
npm run dev         # starts a local n8n with this node loaded

Version history

  • 0.1.0: first release, with the Transcribe operation, five response formats and the Saidwire API credential.

Discussion