I figure this is worth being upfront about because AudioSnag has been built during the exact period where half the internet suddenly seems to have been made by a robot, and putting out anything new now, it's assumed it was all robot made.
AudioSnag wasn’t!!
The original idea, audio processing scripts, workflows and a lot of the weird little visual stuff came from me making tools for my own work. The bitmap visuals, bespoke, no vibe-coded slop. The sausage mascot is a drawing in a notebook. Where I did have to lean on AI, was on frontend UI/UX stuff.
AI inside AudioSnag
99% of AudioSnag does not need generative AI to work.
Some optional features may use online transcription, classification or analysis because those things can genuinely save time.
If using one of those features means sending your audio or information to an outside provider, AudioSnag will tell you before it happens.
The result is also just a result. A transcript can be wrong. An edit suggestion can be stupid. A classification can miss something. You are still the producer.
Most importantly, I do not use your files to train AI models and I do not treat your private work as material for improving AudioSnag. Crazy that needs to be said but yeah.
If you ever think the app is doing something with your material that it hasn’t explained properly, email me: sausage@audiosnag.org.