Skip to content
SPECIFICTION
Get listed

mercuryScribe

mercuryscribe.com

Free plan Open source WebWindowsmacOSLinuxSelf-hosted
Visit mercuryscribe.com
Spec score 8.8 #6 of 25 in Podcast Transcription Software

What it is

mercuryScribe is a locally run web interface for transcribing audio and video, with transcript editing and document export. It uses Faster Whisper for speech transcription and pyannote models to distinguish speakers. Users can edit transcripts in a browser, save changes as VTT files, and generate HTML or DOCX files containing speaker names and timestamps. The FAQ says audio, transcripts, and metadata are not sent to external services; the models are obtained from Hugging Face. Users need a Hugging Face access token and must accept terms for the required models. MAXQDA support is included by default, and custom integrations can be developed. The maker documents Docker, Windows package, and Python package installation options. The software is free and open source under the MIT license for personal, academic, or commercial use, and is listed for Linux, macOS, Windows, and web. The maker describes it as built for researchers and also suitable for journalists, podcasters, organizations, educators, accessibility groups, and professionals. Professional services include installation help, training, workshops, and private server setup with GPUs.

Spec sheet

What it does
mercuryScribe is a browser-based local transcription front end for audio and video files.Read 15 Oct 2026 · github.com
Product
mercuryScribe is a locally run web interface for transcribing audio and video files.Read 3 Oct 2026 · mercuryscribe.com
Intended users
The maker describes it as built for researchers and suitable for journalists, podcasters, organizations, educators, accessibility groups, and professionals.Read 3 Oct 2026 · mercuryscribe.com
Privacy
The maker states that audio, transcripts, and metadata are not sent to external services.Read 15 Oct 2026 · mercuryscribe.com
Compliance
The FAQ says suitability for HIPAA or GDPR depends on the user's implementation and advises consulting their compliance team.Read 3 Oct 2026 · mercuryscribe.com
Support
Professional services include installation help, team training, workshops, and private server setup with GPUs.Read 3 Oct 2026 · mercuryscribe.com
Headquarters
Nashville, Tennessee, USARead 28 Sep 2026 · mercuryscribe.com
Accuracy statement
The FAQ says Faster Whisper Base typically reaches above 80% accuracy on clear English audio, depending on recording quality and language.Read 15 Oct 2026 · mercuryscribe.com
API extensibility
Professional services include API integrations with existing tools, but the public pages do not document a standalone public API.Read 15 Oct 2026 · mercuryscribe.com
Batch processing
The companion command-line interface supports batch transcription from the command line.Read 15 Oct 2026 · github.com
Custom integrations
The provider offers custom integrations with content-management systems, analysis software, and other tools.Read 15 Oct 2026 · mercuryscribe.com
Docker distribution
A multi-architecture Docker web image is published through GitHub Container Registry.Read 15 Oct 2026 · pypi.org
DOCX output
The software generates DOCX files containing speaker and timestamp information.Read 15 Oct 2026 · github.com
Editing
Users can edit transcripts in a browser and save changes to VTT files.Read 3 Oct 2026 · github.com
GPU deployment
Enterprise deployments can use GPU acceleration for initial transcription and cheaper instances for editing and inspection.Read 15 Oct 2026 · mercuryscribe.com
HTML output
Transcription produces an HTML page containing the transcript and synchronized media playback.Read 15 Oct 2026 · github.com
Install options
The maker documents Docker, a Windows package, and Python package installation.Read 3 Oct 2026 · mercuryscribe.com
Integration
MAXQDA support is the default, and custom integrations can be developed.Read 3 Oct 2026 · mercuryscribe.com
Licence
The maker states that mercuryScribe is open source under the MIT licence.Read 15 Oct 2026 · mercuryscribe.com
Limit: model access
Required Hugging Face models may require account creation, access requests, and acceptance of model terms.Read 15 Oct 2026 · github.com
Limit: setup complexity
The FAQ warns that installing Python dependencies can be difficult and recommends Docker for most users.Read 15 Oct 2026 · mercuryscribe.com
Limit: Windows testing
The project documentation says Windows support is not well tested.Read 15 Oct 2026 · pypi.org
Local processing
The application runs on the user’s computer or private server rather than requiring cloud transcription.Read 15 Oct 2026 · mercuryscribe.com
MAXQDA integration
DOCX timestamps are designed to be compatible with MAXQDA, which the maker identifies as the default integration.Read 15 Oct 2026 · mercuryscribe.com
Model access
Users need a Hugging Face access token and must accept terms for the required models.Read 3 Oct 2026 · mercuryscribe.com
Output
The tool generates HTML and DOCX files with speaker names and timestamps.Read 3 Oct 2026 · github.com
Paid services
Enterprise support may include priority support, custom installations, SLAs, and pre-configured solutions; the software itself is free.Read 3 Oct 2026 · mercuryscribe.com
Performance statement
The maker reports approximately real-time processing on tested Intel Linux and Intel or M2 Mac systems.Read 15 Oct 2026 · mercuryscribe.com
Python distribution
The project is installable as the transcribe-with-whisper Python package from PyPI.Read 15 Oct 2026 · pypi.org
Release fact
PyPI lists transcribe-with-whisper version 0.8.3 uploaded on January 30, 2026.Read 15 Oct 2026 · pypi.org
Security and compliance
The maker says local deployment can fit HIPAA or GDPR requirements but advises consulting the organization’s compliance team.Read 15 Oct 2026 · mercuryscribe.com
Self-hosting
Organizations can deploy the software on private cloud, on-premises, DigitalOcean, or AWS infrastructure.Read 15 Oct 2026 · mercuryscribe.com
Speaker diarization
Speaker separation uses the pyannote speaker-diarization-community-1 model.Read 15 Oct 2026 · github.com
Support services
Paid services include guided installation, Docker configuration, training, workshops, priority support, and custom development.Read 15 Oct 2026 · mercuryscribe.com
Supported media
The processing workflow accepts MP4, MOV, MKV, and MP3 inputs, although browser video display is expected to work best with MP4.Read 15 Oct 2026 · github.com
Synchronized playback
Clicking transcript sections can jump the associated video to the matching time.Read 15 Oct 2026 · github.com
System requirement: ffmpeg
The Python installation requires ffmpeg for audio and video conversion.Read 15 Oct 2026 · mercuryscribe.com
System requirement: Hugging Face
Users need a Hugging Face account, authentication token, and access to required models.Read 15 Oct 2026 · github.com
System requirement: Python
The PyPI package requires Python 3.8 or newer.Read 15 Oct 2026 · pypi.org
System requirement: Windows setup
The FAQ recommends WSL for Windows Python-package installation and says dependency setup can be tricky.Read 15 Oct 2026 · mercuryscribe.com
Transcript editing
Users can edit transcripts in a web browser and save edits back to VTT files.Read 15 Oct 2026 · github.com
Transcription
It uses Faster Whisper to transcribe speech and pyannote models to distinguish speakers.Read 3 Oct 2026 · github.com
Transcription engine
Speech recognition uses the Faster Whisper Python library.Read 15 Oct 2026 · github.com
WebVTT workflow
Edited VTT files can be used to regenerate updated HTML and DOCX outputs.Read 15 Oct 2026 · github.com
Windows bundle
Releases include a packaged Windows distribution containing MercuryScribe.exe and a start-server.bat file.Read 15 Oct 2026 · pypi.org

Plans and prices

Free Free No licensing fees Personal, academic, and commercial use · core software free and open-source Read 15 Oct 2026 · mercuryscribe.com

Every mercuryScribe plan and price →

Strong points and limits

Strong points

  • Free and open source under the MIT license.
  • Transcript edits can be saved as VTT.
  • HTML and DOCX exports include speakers and timestamps.
  • FAQ says media and transcripts stay off external services.
  • Docker, Windows, and Python installation options are documented.

Limits

  • Requires a Hugging Face access token.
  • Users must accept terms for required models.
  • HIPAA or GDPR suitability depends on implementation.

Who it suits

It suits researchers and other users who need editable, speaker-labeled transcripts from audio or video. Teams should review their own compliance implementation, as the FAQ directs users to consult their compliance team about HIPAA or GDPR suitability.

Verdict

mercuryScribe describes its transcription workflow, output formats, and local data handling, with no software charge listed. Model access requires a token and acceptance of model terms; compliance suitability is left to the user's implementation.

Where Specifiction places it

Alternatives to mercuryScribeAll alternatives →