Free online audio editor with AI commands

Edit audio without desktop installs

Podcasters, students, and creators often need short audio fixes: cut silence, merge takes, convert formats, or generate a voiceover. Vocix provides those tools in the browser with prompt-based controls and local processing for standard edits.

Upload audio and type prompts like “trim from 0:00 to 0:45”, “merge these audios”, “convert to mp3”, or “transcribe this audio”. Interactive options cover text-to-speech when you need a new voice track from written text.

Audio features

• Trim and cut tracks to precise ranges
• Merge multiple audio files in order
• Convert and compress common formats
• Adjust speed and reduce noise where supported
• Extract audio from video uploads
• AI transcription and text-to-speech with free daily limits

Quality tips

Start with the cleanest recording you can. Denoise helps mild background hiss but will not fix severe distortion. For voiceovers, write the script first, generate TTS, then trim and merge into your final track.

Always listen through headphones before publishing. Browser editors are convenient, but your ears catch clicks and clipped peaks that a waveform alone might hide.

Podcast and voiceover mini-workflows

Solo podcast polish: upload the take → trim dead air → optional denoise for mild hiss → compress/convert to MP3 → done. Interview polish: trim each side if recorded separately, merge in conversational order, then listen for double-breaths at the join.

Voiceover from script: write offline → Text to Speech for a draft bed → trim TTS edges → merge with music bed carefully so voice sits forward. TTS uses daily quota; local trims afterward do not.

From video lectures: extract audio first, then edit the lighter file. You avoid dragging picture frames through every experiment. Students often need a five-minute excerpt—trim locally, then decide whether transcription is worth a credit.

Denoise, convert, and format expectations

Denoise helps steady hiss and light HVAC rumble. It will not repair clipping, wind blasts, or crushed voice notes. If the waveform is a solid brick of distortion, re-record.

Convert when a platform demands MP3 or when a huge WAV wastes upload quotas on later AI steps. Keep a master copy before aggressive compression. For transcription, clearer always beats smaller—do not crush bitrate so hard that consonants vanish.

Speed changes are for pacing tutorials or compressing silence-heavy meetings. Announce sped speech in educational contexts so listeners are not surprised. Pair speed changes with a quick headphone pass to catch artifacts at consonants.

• Trim silence before you denoise long files.
• Headphone-check every export for clicks at edit points.
• Name versions: _raw, _trim, _master to avoid overwriting.
• Spend TTS quota on final scripts, not first drafts.

Audio troubleshooting

Merge order wrong: re-upload in the correct sequence; the first file should be the open. Speed changes making chipmunk speech: you asked for too much—try a milder factor or cut duration instead. Transcription refusing noisy files: trim to a clean minute and specify language.

If local FFmpeg fails on a huge lossless project, bounce a 48 kHz AAC/MP3 proxy for editing, then apply the same cut points on a desktop DAW for the final master when bit-perfect delivery matters.

Account required for the editor means quotas and abuse controls stay attached to a real session; guides stay readable without login. If AI voice generation hits the daily cap, continue mechanical edits until midnight UTC reset.

Example repair sequence for a noisy field interview: trim to the usable answers → light denoise → convert to MP3 → optional transcription with the correct language. Stop after denoise if the file is still unintelligible; more processing will not invent missing consonants.