About Waveform AI
We think everyone should have studio-grade audio tools, without a studio budget.
Our story
We started as a side project — a way to remove vocals from a favorite track to practice karaoke. It turned out a lot of people wanted the same thing, for very different reasons: podcasters cleaning up dialogue, musicians sampling old records, indie developers who needed a narrator but no budget for one.
Today the same underlying models power vocal removal, voice cloning, stem separation, and music generation — one platform, built by a small team who use these tools ourselves every day.
Purpose-built audio models, not a general-purpose wrapper.
What we value
Your audio stays yours
Uploads are processed and deleted — we never train on your content or claim ownership of what you create.
Fast, not just accurate
Most jobs finish in under 10 seconds. We optimize for creators who are mid-workflow, not waiting on a render queue.
Built with our users
Most of our roadmap comes directly from creator feedback — we ship what people actually ask for.