LaunchAI ModelsOctober 8, 2026

Cactus Compute releases Whistle, a 16.9 MB on-device speech-to-text model

Read original source →cactuscompute.com

Whistle is a single 16.9 MB file that runs on CPU with no dependencies, supporting English, German, French, Spanish, Italian, Dutch and Polish. It mostly beats Whisper base with 9x less file size and 6x speed, and also outputs word timestamps and speech embeddings.

How this story unfolded

4 days · 1 report · 3 community posts · from Oct 5

  1. Oct 5
  2. Oct 8
  3. Oct 9

More stories today

Open the live feed