AudioForges is a browser studio for taking songs apart. Drop a song and get clean vocals, an instrumental, or separate drums, bass, guitar, piano and other stems in about a minute. Listen to every stem in Forge Mixer before you download, then turn the result into a backing track, a karaoke version, MIDI or engraved sheet music.
It’s built by a solo music producer for singers, cover bands, worship teams, DJs and producers. No install and no plugins. You can start free, pay per song, or get the Studio Pass for regular use.

Hey Fazier 👋 I’m Dipak, a music producer from Kathmandu. AudioForges started as a small tool I built for my own sessions. I kept needing to pull an acapella, find the key of a sample, or make a backing track to practice over, and every tool I tried either wanted a subscription before I could hear anything or gave me muddy stems. So I built my own. Today it’s a full browser studio: drop a song and you get vocals, instrumental or up to 6 stems, and you can play them all in Forge Mixer before downloading. From there you can turn the song into a backing track in your own key, a Stage Pack for live shows, MIDI you can edit in the piano roll, or sheet music you can play along with. It’s built and run by just me, so every bit of feedback reaches the person writing the code. I’d love to hear: How the separation sounds on your own songs What’s missing for your workflow (singing, DJing, producing, live) You can try it free, no install needed. Thanks for checking it out!

I build tools for YouTube creators, and the backing-track maker is the feature that stands out for that audience — being able to take a song, drop a part, and shift key and tempo for a cover video without leaving the browser is genuinely useful. Letting people listen to every stem in the Forge Mixer before downloading is also a smart call; that is exactly where other splitters waste your time and credits. One question: when you export sheet music, does it handle multi-instrument scores or is it a single combined staff? Congrats on the launch!
The combination of stem separation plus MIDI and sheet music in one browser workflow is what stands out — most tools stop at stems. I'm curious how the audio-to-MIDI handles dense piano parts or heavily reverberant vocals, since that's where transcription usually falls apart. Does the Forge Mixer let you export a rebalanced mix as well as individual stems? The lead vs backing vocal split could be a big deal for cover bands making backing tracks in their own key.

Hey Fazier 👋 I’m Dipak, a music producer from Kathmandu. AudioForges started as a small tool I built for my own sessions. I kept needing to pull an acapella, find the key of a sample, or make a backing track to practice over, and every tool I tried either wanted a subscription before I could hear anything or gave me muddy stems. So I built my own. Today it’s a full browser studio: drop a song and you get vocals, instrumental or up to 6 stems, and you can play them all in Forge Mixer before downloading. From there you can turn the song into a backing track in your own key, a Stage Pack for live shows, MIDI you can edit in the piano roll, or sheet music you can play along with. It’s built and run by just me, so every bit of feedback reaches the person writing the code. I’d love to hear: How the separation sounds on your own songs What’s missing for your workflow (singing, DJing, producing, live) You can try it free, no install needed. Thanks for checking it out!

I build tools for YouTube creators, and the backing-track maker is the feature that stands out for that audience — being able to take a song, drop a part, and shift key and tempo for a cover video without leaving the browser is genuinely useful. Letting people listen to every stem in the Forge Mixer before downloading is also a smart call; that is exactly where other splitters waste your time and credits. One question: when you export sheet music, does it handle multi-instrument scores or is it a single combined staff? Congrats on the launch!
The combination of stem separation plus MIDI and sheet music in one browser workflow is what stands out — most tools stop at stems. I'm curious how the audio-to-MIDI handles dense piano parts or heavily reverberant vocals, since that's where transcription usually falls apart. Does the Forge Mixer let you export a rebalanced mix as well as individual stems? The lead vs backing vocal split could be a big deal for cover bands making backing tracks in their own key.
Find your next favorite product or submit your own. Made by @FalakDigital.
Copyright ©2026. All Rights Reserved