Jan 29 · 4 min read · Voice input is becoming table stakes for modern web apps. Users expect to tap a mic button and speak instead of typing — especially on mobile. Here is how I added real-time voice transcription to a React form using OpenAI Whisper. The Architecture Th...
Join discussionNov 24, 2025 · 3 min read · 📝 Quick Summary: noScribe is a free and open-source AI-based software for automated audio transcription, primarily designed for qualitative social research and journalistic use. It leverages Whisper and pyannote for transcription and speaker identif...
Join discussionNov 7, 2025 · 4 min read · 前言 本文檔對比三種使用 Whisper 進行語音轉文字的方式，包含速度、成本與實際使用建議。測試音訊為 56 分 27 秒（約 52 MB）的訪談錄音。 【 總結 】 追求速度與性價比平衡：RunPod RTX 4090 速度最快（75 秒） 成本合理（$0.023/次） 無文件限制 適合：中高頻使用（>50 次/天） 零部署需求：Fireworks V3 Turbo API 呼叫即可使用 成本中等（$0.051/次） 速度中等（100 秒） 適合：低頻使用（<50 次/...
Join discussionNov 5, 2025 · 2 min read · 📖 What Are Speech Models? Speech models are advanced AI systems designed to convert spoken language into written text (speech-to-text) and sometimes even the reverse (text-to-speech).They are the foundation of voice assistants, transcription tools, ...
Join discussion
Oct 8, 2025 · 5 min read · TL;DR: Built a privacy-first voice-to-text app for Linux that works completely offline. Types directly into any application. Open-source, production-ready, comprehensive test coverage. The Problem I wanted to create a Linux voice typing app that: Wo...
Join discussion
Oct 9, 2025 · 5 min read · Czy kiedykolwiek marzyłeś o tym, żeby automatycznie transkrybować spotkania i tłumaczyć je na swój język? W tym artykule pokażę Ci, jak stworzyć własny system AI do transkrypcji i tłumaczenia audio używając Python i modeli Hugging Face. 🎯 Co to jest...
Join discussion
Jul 7, 2025 · 5 min read · Whisper is computationally intensive, and running it efficiently at scale or on constrained devices is a real engineering challenge. To make Whisper more suitable for production deployment, I explored two independent and hardware-aware optimization s...
Join discussion
Jul 7, 2025 · 5 min read · The Whisper model by OpenAI is a powerful multilingual, multitask ASR system. However, this generality can become a liability when adapting Whisper to a specific language and task. For example, transcribing Azerbaijani speech without translation or l...
Join discussion
Jun 18, 2025 · 5 min read · Introduction Creating a local speech-to-speech translator is now more accessible than ever, thanks to advancements in open-source speech recognition, translation, and text-to-speech technologies. In this blog, I'll walk you through how I built a full...
Join discussion
Jun 7, 2025 · 4 min read · Introduction There are several online tools that can generate transcripts simply by providing a YouTube video URL. These work well for public videos, making them a convenient solution in most cases. However, when it came to member-only videos, I quic...
Join discussion