A developer built ttsproof, an open-source QA framework for text-to-speech systems that separates structural defects from pronunciation errors, addressing the shortcomings of Word Error Rate metrics. …
A developer building a phone assistant that transcribes work calls has chosen to run speech-to-text locally with faster-whisper instead of using cloud APIs, citing data privacy concerns. The local set…
A developer's RAG pipeline for MockEvalio failed on its first day when the embedding endpoint returned a 404, prompting a switch from Groq's nomic-embed-text-v1.5 to OpenAI's text-embedding-3-small. T…
A developer known as C.Studva has open-sourced Jarvis CLS, a local-first AI assistant for macOS and Linux under the MIT license. The assistant runs entirely on-device with no cloud dependency, using f…
A developer built a fully local voice-to-code pipeline using faster-whisper for transcription and Claude Code's /voice command, ensuring no data leaves the machine. The setup, which runs on CPU with t…
A solo developer rebuilt an AI transcription pipeline on AWS using Spot Instances, SQS queue-based autoscaling, and Whisper Large-v3 Turbo, cutting processing costs below $0.10 per hour—cheaper than t…
A new open-source MCP server called talkthrough-mcp, released under the MIT license, lets developers turn narrated screen recordings into structured data for AI agents without any cloud dependency. Th…
A developer built claude-real-video, an MIT-licensed tool that converts video files into formats LLMs can genuinely process, addressing the gap where models like ChatGPT and Claude fail to handle actu…
A free AI transcription tool using faster-whisper on Google Colab's T4 GPU was released, allowing users to upload audio files, generate .srt subtitle files, and play them back with SRT Player.…
A developer built a script that takes an English video and produces a Spanish-narrated version with a cloned voice of the original speaker, using faster-whisper for transcription, an LLM for translati…
OpenAI Whisper has 12 unique checkpoint files, but the Python package exposes 14 local model names due to aliases for 'large' and 'turbo'. Additional confusion arises from runtimes and converted forma…
A Quebec-based developer built Tania Dictée, a Windows-only voice dictation tool that handles code-switching between French and English technical terms. The tool processes each phrase individually usi…
A developer built LocalScribe PRO, a free, offline, and private meeting transcription tool that runs entirely on a laptop. The tool uses Python, Ollama for local AI, and faster-whisper for speech-to-t…
A developer built a Windows dictation tool optimized for Quebec French and code-switching between French and English, using Whisper models locally. The tool, which costs $9 CAD, was created after find…
Nemorize has published a learning roadmap titled "Building Real-Time Voice Agents from Scratch," covering the full pipeline from audio fundamentals and speech detection to ASR, LLM streaming, and TTS.…