# Speaker Separation for Long Recordings

AudioPod splits long meetings, lectures and panels into one track per speaker with its AudioDiarize engine: upload the file or paste a link and each voice comes back on its own full-length track. Jobs run up to 30 minutes on free Basic and 2 hours on Creator, at 330 credits per minute.

Canonical URL: https://audiopod.ai/speaker-separation/long-recordings
Last updated: 2026-10-16

## Facts

| Fact | Detail |
| --- | --- |
| Longest recording per job | Basic 30 minutes · Creator 2 hours · Pro 5 hours · Studio 10 hours · Enterprise: no cap |
| Largest upload | Basic 500 MB · Creator 2 GB · Pro 4 GB · Studio and Enterprise 4.99 GB |
| Pasted links | Limited by length only, not file size |
| Files per batch | Up to 3 at once in the dashboard, each its own job |
| Video files | MP4, MOV, AVI, MKV, WEBM are accepted; the audio is separated |
| One hour of audio | 19,800 credits (about $2.64 pay-as-you-go) |
| What the credits cover | Basic: about 15.1 minutes a month · Creator: about 10 hours a month |
| Speakers | Auto detect, or 2–10 speakers |
| Each track | As long as the original, silent where that person is not talking, so tracks line up in any editor |
| Same file twice | Resubmitting an identical job while the first is still running (started under 3 hours ago) returns that job instead of charging again |

## How to do it

1. Open Speaker Separation.
2. Upload up to 3 recordings, or paste a link.
3. Leave Auto detect on, or set the number of speakers for a known panel.
4. Check the quote for the total length and start.
5. Download each speaker's track as WAV or MP3.

## What long recordings need

- Speech that falls between detected turns is not lost: it goes to the speaker nearest in time, while room tone and long silences stay silent.
- A speaker with less than 0.5 seconds of speech, such as someone who only says hello, gets no track.
- If the count comes out wrong, set the speaker count yourself (2–10) instead of Auto detect.
- Credits are reserved when the job starts and refunded if it fails.

## If the recording is longer than your plan allows

- Upgrade for a longer cap: 2 hours on Creator, 5 hours on Pro, 10 hours on Studio.
- Or split the file into parts at natural breaks and run each part; speaker numbers are per job, so match them up across parts.
- A file that is too large but short enough can come in as a link instead of an upload.

## FAQ

**How long can a recording be for speaker separation?**
It depends on the plan: Basic 30 minutes · Creator 2 hours · Pro 5 hours · Studio 10 hours · Enterprise: no cap. Uploads are capped at Basic 500 MB · Creator 2 GB · Pro 4 GB · Studio and Enterprise 4.99 GB. A pasted link is limited by length only.

**How much does separating a one-hour meeting cost?**
19,800 credits, at 330 credits per minute. That is about $2.64 pay-as-you-go, and Creator's 200,000 monthly credits cover about 10 hours. See /pricing for plan prices.

**Can I upload a video of the meeting?**
Yes. MP4, MOV, AVI, MKV, WEBM files are accepted and their audio is separated by speaker. You can also paste a link to the recording.

**How many speakers can it find?**
In the dashboard you choose Auto detect, or 2–10 speakers. For large panels where automatic detection merges two similar voices, setting the count explicitly gives the best result.

**Do the speaker tracks line up with the original?**
Yes. Every track is as long as the original recording and silent wherever that speaker is not talking, so you can drop all of them into an editor at zero and they stay in sync.

## Related

- [Speaker Separation](https://audiopod.ai/features/speaker-separation) — The tool, the free demo and the speaker guides.
- [Overlapping voices](https://audiopod.ai/speaker-separation/overlapping-voices) — What happens when two people talk at once.
- [Speaker-labeled transcript](https://audiopod.ai/speaker-separation/speaker-labeled-transcript) — A transcript with speaker names, at no extra cost.
- [Speech to Text](https://audiopod.ai/features/speech-to-text) — Transcription and subtitles when you need text, not tracks.

## Pricing

Plan prices: [pricing](https://audiopod.ai/pricing). Start: [Split a long recording](https://audiopod.ai/auth/signup?callbackUrl=%2Fdashboard%2Ftools%2Fspeaker-separation).