EmpirioLabs AI Transcribe

Transcribe audio or video files to text with EmpirioLabs AI, from an https URL or a local file. Use when the user asks for a transcript of a recording, meeting, interview, podcast, voice note, or video, wants speaker-labeled text, or needs the spoken words as text for notes, summaries, or captions. Runs a bundled Python script that picks a transcription model from the live catalog (or the one you name) and waits for the result.

EmpirioLabs AI transcription

Turn speech in an audio or video file into text with the EmpirioLabs AI transcription API. Transcription runs as a job: the bundled script submits the file, waits, and prints the transcript.

Setup

1. Create an API key at https://platform.empiriolabs.ai/dashboard/api-keys. Keys start with sk-empiriolabs-.
2. Make the key available as EMPIRIOLABS_API_KEY in the environment where commands run. Never paste the key into a prompt or a file in the project.
3. Python 3.8 or later. There is nothing to install.

Script paths below are relative to this skill's folder. On Windows, use python instead of python3.

Transcribe a file

bash

python3 scripts/transcribe.py https://example.com/interview.mp3
python3 scripts/transcribe.py ./recordings/meeting.m4a --out ./notes/meeting.txt

The source can be an https URL to an audio or video file, or a local file, which the script uploads. Without --out, the transcript is printed to stdout. With --out, it is saved (a .json path saves the full result, including timing data where the model returns it) and a JSON summary is printed:

json

{
  "model": "whisper-large-v3-turbo",
  "job_id": "...",
  "file": "/home/me/project/notes/meeting.txt",
  "characters": 5321
}

Speakers, languages, and full results

bash

python3 scripts/transcribe.py https://example.com/panel.mp4 --model deepgram-nova-3 --diarize --out ./panel.json
python3 scripts/transcribe.py ./voice-note.ogg --language es
python3 scripts/transcribe.py ./call.wav --json        # full result with segments and words

Long recordings

bash

python3 scripts/transcribe.py https://example.com/webinar.mp3 --no-wait   # prints the job id
python3 scripts/transcribe.py --job-id JOB_ID --out ./webinar.txt           # waits, then saves

If --timeout (default 1800 seconds) runs out, the job keeps running. The script exits with code 3 and prints the --job-id command that resumes it.

Options

Option	Meaning
--model ID	Use a specific model, for example whisper-large-v3-turbo or deepgram-nova-3.
--language CODE	Spoken language hint, for example en, on models that accept it.
--diarize	Label speakers, on models that support it.
--param key=value	Any other model parameter, for example --param smart_format=true.
--out PATH	Save the transcript (.txt) or the full result (.json).
--json	Print the full result as JSON instead of plain text.
--no-wait / --job-id ID	Submit and return, then resume later.

Choose a model and check prices

bash

python3 scripts/transcribe.py --list-models

Transcription is billed by audio length at the model's listed price. * marks the model used when --model is omitted. For a model's options, use the empiriolabs-models skill (models.py --show MODEL_ID).

Exit codes and errors

• 0 success, 1 the request failed (the message says why), 2 invalid arguments or a missing file, 3 still running when --timeout ran out.
• HTTP 401: the key is missing or invalid. HTTP 402: add credits at https://platform.empiriolabs.ai/dashboard/billing. HTTP 413: the upload is too large; pass an https URL to the file instead.

Other ways to connect

The same models are available through the EmpirioLabs AI remote MCP server at https://mcp.empiriolabs.ai/mcp (OAuth sign-in, or the API key as a bearer token). Setup: https://docs.empiriolabs.ai/mcp. API reference: https://docs.empiriolabs.ai.