Show HN: Superwhisper S1 mini – Text and tone normalizer for speech-to-text
Category: ai-ml
Tags: llm-inference, gguf, speech-to-text, text-normalization, cpp
Score: 7.8/10 (Innovation: 6, Technical: 9, Documentation: 8, Utility: 8)
The submission points to a GGUF model for text and tone normalization of speech-to-text output, built on top of llama.cpp. The underlying llama.cpp project is a highly optimized C/C++ LLM inference engine supporting many hardware backends, which makes running such small models on laptops feasible. The model itself addresses a practical niche: cleaning raw ASR transcripts into well-formatted written text.
Target audience: machine learning engineers, backend developers, speech application developers
Repository: https://huggingface.co/superwhisper/s1-mini-GGUF · C++ · MIT · 131946 stars
View on Hacker News