Cactus Compute

Developers are discussing Whistle, a compact speech-to-text model released by Cactus Compute that fits in just 16.9 MB. The small footprint makes on-device transcription practical for mobile and embedded applications where larger models are impractical. Commenters are weighing its accuracy against heavier alternatives like Whisper, and debating how far small automatic speech recognition models can now go.

Developers are highlighting Whistle, an open-source speech-to-text model that fits in just 16.9 MB, detailed in a post by Cactus Compute. The model's small footprint makes it practical for running transcription locally on constrained devices. Tech communities are sharing the link and discussing its potential for on-device AI.