Isolate Voice From Two MP3 Files

I have two MP3 recordings, each running just under 50 minutes, where spoken content sits on top of background music. I simply need the speech extracted and delivered as clean, stand-alone audio.

I’ll be using the result for further voice-over work, so the vocal track must retain its natural tone and timing. Please don’t add noise reduction, EQ, reverb, or any other processing—straight separation is all that’s required.

Deliverables
• Two MP3 files (voice-only, same length as originals, minimal artefacts)

If you’re comfortable with tools such as iZotope RX, Adobe Audition, or any reliable AI source-separation solution and can turn this around quickly, I’d love to hear from you.