[jordimas]

Project

whisper-ctranslate2

Adds an OpenAI Whisper-compatible CLI built on faster-whisper, with CPU/GPU support, live transcription, and optional speaker diarization.

Checklist

• I read CONTRIBUTING.md
• One project per PR
• PR title format: Add project-name
• Entry follows the required format
• Display name is the PyPI package name
• Placed in the existing Speech category
• Active, documented, established, and distributed as a stable package

Which Tier

• Obvious choice
• Challenger

Proposed as a challenger to the original Whisper CLI for users who want familiar command-line options with faster-whisper inference. It complements the existing faster-whisper library entry with a ready-to-use CLI.

Evidence checked on October 30, 2026:

• PyPI Stats: 149,898 downloads in the last month and 64,592 in the last week. These are download counts, not unique users or proof of sustained growth.
• PyPI release history: distributed since 2023; latest release 0.5.8 on October 5, 2026.
• Community use: a user describes using it for local transcription.
• README: installation, CLI examples, Docker deployment, and feature documentation.

Whether that adoption merits a challenger slot is an editorial judgment. No performance claim is added to the list. I found no existing entry or prior submission under whisper-ctranslate2 or Softcatala.

Displacement

Not required: Speech currently has four entries; this would be the fifth.