Convert by ngram
Strip the audio from any video. Clean MP3 or WAV, in seconds.
Drop the video, paste a link, or pull from YouTube. ngram extracts the audio track without re-encoding, hands you an MP3 or WAV ready for podcasts, transcription, and editing.
This conversion isn't available yet. Browse all workflows to find one that's live.
Trusted by teams at
How it works
Four steps. About five seconds of waiting.
No timeline, no codec wrangling, no FFmpeg flags. Drop a video, pick MP3 or WAV, download the clean audio track.
Drop the video in
An MP4, MOV, WebM, MKV, AVI, or M4V file, a public link, or a YouTube / Loom URL. We sniff format, length, and audio stream automatically.
We pull the audio track
The audio stream is demuxed straight from the container. No re-encode by default, so the source quality survives intact — ideal before transcription or podcast editing.
Pick MP3 or WAV
MP3 for podcasts and shareable transcripts, WAV for editing in Logic, Audition, or Descript. Bitrate, sample rate, and channel layout exposed when you need them.
Ship the audio
Download the MP3 or WAV, or hand it straight to ngram for transcription, audiograms, or a new audio-to-video project. Source files auto-delete after 24h.
Output controls
Smart defaults. Real knobs when you need them.
Lossless demux by default
We copy the audio stream as-is when the source codec maps cleanly to MP3 or WAV. No re-encode artifacts, no quiet bitrate loss on the way out.
MP3 or WAV per export
MP3 at 128, 192, 256, or 320 kbps for podcasts, social, and transcription. WAV 16-bit or 24-bit when you need editor-ready masters.
Trim to the podcast segment
Scrub the video, lock the in/out point to the second, and pull only the segment you need. Perfect for cutting an audiogram out of a long interview.
Optional cleanup pass
Flip on dialogue isolation and background noise removal before the MP3 lands. Useful for Zoom and Riverside recordings with room hum or fan noise.
Loudness normalization
Target -16 LUFS for podcast platforms or -14 LUFS for YouTube and Spotify. Levels balance across speakers without manual gain riding.
Batch a folder of recordings
Drop a folder of Zoom or Loom MP4s, walk away. Parallel demux, identical export settings, one ZIP back with every track.
Advanced flags exposed
Mono downmix for transcription pipelines, channel split for multi-track recordings, sample rate forced to 16 kHz when Whisper expects it.
Source files gone in 24h
Processed in-region, encrypted at rest, never used to train models. in-region processingtive.
The rest of ngram
Pulling audio out is the easy bit. The next steps live here.
Captions
Once the audio is out, transcription takes 30 seconds. Burned-in captions on the source video, or an.srt next to the MP3 for podcast show notes.
Learn moreTranslation
Send the extracted MP3 through translation to spin a Spanish or French version of the same interview, voiceover, or webinar audio.
Learn moreAI Voiceover
Pull the script out of the extracted audio, then re-record the voiceover with an ElevenLabs voice when you need a clean studio take instead of the Zoom original.
Learn moreMusic
Layer licensed beds under the extracted dialogue track. Auto-ducking keeps the voice on top in audiograms and podcast trailers.
Learn moreMulti-format Export
Same source video, multiple outputs in one render: MP3 for the podcast feed, MP4 with captions for LinkedIn, audiogram square for Twitter.
Learn moreEnterprise Integrations
Drop the MP3 into a Zapier or n8n flow on completion — ship it to Descript, Otter, AssemblyAI, or your podcast host without a human in the loop.
Learn moreUse cases
Where pulling clean audio out of a video earns its keep.
Webinar audio becomes a podcast episode
Extract the audio track from a 60-minute webinar MP4, run it through cleanup, and publish it as the week's podcast — same content, second channel.
See use caseAudiograms from interview footage
Pull the audio from a long-form interview, cut the soundbite to 60 seconds, and pair it with a static brand frame for LinkedIn and X.
See use caseTranscripts of recorded demos
Extract the audio from a recorded sales call, feed the MP3 to your transcription tool, and turn the result into a follow-up email or CRM note.
See use caseSearchable archives of internal calls
Pull MP3s out of recorded Zoom or Meet sessions, archive them next to written notes, and surface the spoken decisions in search.
See use caseAudio-only lessons for the course library
Pull the narration out of recorded lecture videos and ship audio-only lessons alongside the video versions for commute and gym listeners.
See use caseVoice-only help responses
Strip the audio out of a recorded screen-share fix, send the MP3 as a voice note inside the ticket. Faster than a Loom, easier than typing.
See use caseReuse narration in a new ngram project
Pull the voiceover from a product demo, import it into a new ngram project as audio input, and pair it with updated screenshots.
See use caseRead-aloud newsletter editions
Strip the audio off a recorded weekly recap video and embed the MP3 in the newsletter as the read-aloud edition for subscribers.
See use caseQuote audio pulled from Zoom calls
Extract the cleanest 30-second quote out of a customer Zoom recording and drop the WAV into a brand video edit.
See use caseOther converters
Coming from something else? There's a converter for that.
Same engine, different input. Video to Audio is one of 16+ converters that all share the same processing pipeline, security model, and brand kit.
The other direction. Pair the MP3 you just pulled out with branded waveforms, captions, and B-roll for an audiogram.
Open converterAn hour of recording becomes 20+ short clips, auto-cut on the highest-attention moments — audio or video out.
Open converterWhen the highlight moment needs to autoplay inline instead of waiting for the audio to load.
Open converterTools that pair with this converter
Once you have the audio, here's what comes next.
Working with the extracted audio
Transcribe, clean, and remix the MP3 you just pulled
Audio to Text
Run the extracted MP3 through transcription for podcast show notes, sales call summaries, or accessible transcripts.
Open toolBackground Noise Remover
Strip room hum, keyboard clicks, and HVAC out of the MP3 before it lands in a podcast feed.
Open toolVoice Dubber
Take the extracted dialogue track and re-voice it in a different language while keeping the timing intact.
Open toolAI Voice Generator
Use the extracted script to re-record cleaner voiceover when the source audio is too noisy to ship.
Open toolGenerating from scratch
When the source video isn't ready yet
Text to Speech Video
From paragraph to narrated clip. Skip the recording step entirely when you only need the voiceover.
Open toolAudio to Video
Pair the extracted MP3 with a branded waveform, captions, and B-roll for an audiogram in one render.
Open toolAI Video Generator
Prompt to finished cut. Drop the extracted audio in for a fully branded video around your original voiceover.
Open toolEditing the source video
Take care of the video before extracting audio
Video Editor
Trim or split the source video first so you only extract audio from the segment that matters.
Open toolVideo Cutter
Cut by transcript, not timeline — isolate the exact spoken segment before pulling the audio.
Open toolVideo Compressor
Shrink a 4 GB recording below the 500 MB upload limit before sending it through the audio extractor.
Open toolVideo Translator
Translate the video first, then pull a localized audio track per language for podcast distribution.
Open toolBuilt for teams
Who reaches for video-to-audio in your company?
Content Creators
Pull podcast-ready MP3s out of interview footage. Same recording, two channels: YouTube video and audio podcast.
See workflowsProduct Marketing
Repurpose webinar and event recordings into audio-only content for podcast feeds and audiograms.
See workflowsSales Enablement
Extract audio from recorded demos and discovery calls, transcribe for follow-up emails and CRM notes.
See workflowsCustomer Success
Pull MP3s out of QBR recordings and customer education videos for searchable account archives.
See workflowsDeveloper Relations
Extract conference talk audio for podcast cross-posts and transcript-driven blog posts.
See workflowsEducators
Ship audio-only versions of lecture videos for students who learn on commutes or during workouts.
See workflowsSupport Teams
Strip the audio off a recorded fix and send it as a voice note inside the ticket — quicker than a fresh Loom.
See workflowsHR & Internal Comms
Pull audio from all-hands recordings for the asynchronous catchup feed remote teammates listen to.
See workflowsIntegrations
Triggers, not logos. Plug video-to-audio into your real workflow.
Every integration ships with a working template. Start from one, or wire your own with the REST API and webhooks.
whenA Zoom or Riverside recording lands in your workspace
thenPull the audio as MP3 and drop it in your podcast host's inbox
whenClaude or ChatGPT calls the audio extractor on a video URL
thenReturn the MP3 plus a transcript-ready download link
whenA self-hosted flow drops a recorded interview MP4 on S3
thenExtract the audio inside your VPC and ship the WAV onward
whenA HubSpot deal moves to 'Demo recorded'
thenPull the audio from the linked recording and attach the MP3 to the deal
whenYou hit 'Extract audio' on a YouTube, Vimeo, or Loom tab
thenGet an MP3 back in a new tab in about five seconds
whenA YouTube URL is pasted into the extractor
thenFetch the video and demux the audio track straight to MP3
How it compares
If you've been using something else for this.
Online extractors and desktop tools both work, mostly. CloudConvert, VEED, and Kapwing handle the basic demux. ngram does that, plus everything that happens before and after — cleanup, transcription, audiograms, branded export.
| Feature | ngram | CloudConvert | VEED | Kapwing |
|---|---|---|---|---|
| Max input file size | 500 MB | 1 GB (paid) | 250 MB | 1 GB (paid) |
| Output formats | MP3 + WAV with bitrate control | MP3, WAV, AAC, FLAC | MP3 only | MP3 only |
| Lossless audio demux | Default — no re-encode when codecs match | Re-encodes by default | Re-encodes | Re-encodes |
| Trim before export | Sub-second, scrub by thumbnail | Whole-second | Whole-second | Whole-second |
| URL ingest (YouTube, Loom) | Native | URL upload only | URL upload only | URL upload only |
| Built-in noise cleanup | Background noise removal + dialogue isolation | — | Separate tool | Separate tool |
| Loudness normalization | LUFS-targeted per platform | — | Basic gain | Basic gain |
| Batch processing | Folder upload, parallel | Sequential queue | One at a time | One at a time |
| API + webhook | REST, MCP, n8n, Zapier | API only | — | API only |
| Files auto-deleted | 24h | 24h | Variable | Variable |
FAQ
Common questions about video to audio
Still curious?
Video → Audio
Ready to pull clean audio out of any video?
Drop the source, pick MP3 or WAV, ship the file. About five seconds, no install.