Spet, free no-signup transcription that's good at non-English audio (spet.app)

Spet has launched a free, no-signup transcription service that excels in translating non-English audio into text. Utilizing OpenAI's Whisper large-v3 model, Spet distinguishes itself from many other free transcription tools that primarily cater to English. It supports over 90 languages, delivering impressive accuracy across languages such as Ukrainian, Russian, Turkish, and several Western European languages. Users can easily upload audio or video files, including Telegram voice messages, and receive their transcriptions without any data retention, as all uploaded audio is deleted post-processing. This tool is significant for the AI/ML community because it democratizes access to high-quality transcription services, particularly for diverse language speakers who often encounter inaccuracies in mainstream tools. Spet’s direct handling of Telegram voice messages without requiring conversion adds a layer of convenience for users relying on popular messaging platforms. With support for multiple audio formats, users can transcribe files of up to 60 minutes and 100 MB, making it a versatile option for a wide range of applications in content creation, accessibility, and multilingual communication.