VoiceBot
- 4 Devlogs
- 5 Total hours
A Telegram bot that transcribes voice messages or videos via Groq API.
A Telegram bot that transcribes voice messages or videos via Groq API.
Today I got the bot completely ready for shipping! I took a screenshot of it in action, beefed up the README, and did a final test of the CI/CD pipeline on my Nest server.
Now, all I have to do is push the code, and my Nest server handles the rest automatically :D
I finally fully deployed my VoiceBot on the Nest server along with the local telegram-api-server, which lets me process videos up to 2GB! The database is also fully up and running now, starting to collect data. Basically, anyone can now look up @egvoicebot on Telegram, add it to their group, and it’ll just work :D
(Btw, Groq gives me an 8-hour audio processing limit for free!)
Today I decided to add the official telegram-api-server to my Nest project. Why? Because this way my bot will be able to handle videos up to 2GB instead of the generic 20MB limit.
Honestly, this project is teaching me how to work with docker-compose files way better than any of my previous projects :)
I decided to build VoiceBot — a Telegram bot that extracts and transcribes audio from video files, video notes , and standard voice messages.
It’s a pure QoL tool built for that one painfully relatable situation: when your friend refuses to use their keyboard and bombards you with 3-minute voice notes while you are stuck in a meeting, a loud train, or just plain hate listening to audio.
Funnily enough, Telegram actually has this feature built-in… but only if you pay for Telegram Premium :D