Podcast File Size: Bitrate, Mono vs Stereo and Host Limits

A one-hour podcast at 128 kbps is about 58 MB, and the same hour at 64 kbps is about 29 MB. The arithmetic is bitrate times duration: kilobits per second times 3,600 seconds, divided by 8, in decimal megabytes.

Hosts quote nearly the same numbers. beehiiv puts 60 minutes of 128 kbps mono MP3 at about 55 MB, and Castos says about 60 MB for a one-hour episode at 128 kbps. Our measured fixture, with its method, is in compress a podcast episode without wrecking the voice.

For an RSS feed, Apple lists 64 to 128 kbps for mono and 128 to 256 kbps for stereo at 44.1 or 48 kHz. Hosts such as Transistor and Castos recommend around 128 kbps. Spotify supports MP3 only between 96 and 320 kbps.

The tension is Spotify against Apple. A 64 kbps mono file sits inside Apple’s range but below Spotify’s stated 96 kbps floor, so a show published to both is safest at 96 kbps or higher in MP3. Apple also asks for AAC where possible: for the same bitrate, it says AAC sounds better. Codec behavior at low rates is covered in audio bitrate, explained.

Mono is smaller only if you also lower the bitrate. At a fixed 64 kbps, a stereo MP3 and its mono downmix were both 514,133 bytes in our test (encoded with ffmpeg, not Smol), so the bitrate you set decides the size, not the channel count.

That result comes from the measurements in the podcast audio post, run on a 64-second two-voice speech fixture on 26 September 2026. The practical rule for mono vs stereo podcast files: use mono for one microphone or for voices that were never placed left and right, since stereo then stores the same signal twice. Use stereo when the show has music beds, ambience or deliberate panning. beehiiv suggests 128 kbps mono for voice and 192 kbps stereo with music.

It is the largest single episode file a host accepts, and it varies widely: 500 MB at beehiiv, 4 GB at Castos, 2,000 MB at Transistor. Several hosts also recommend staying well under their hard cap.

Caps are only half the bill. Some plans meter storage or monthly upload volume, and others meter bandwidth: Buzzsprout includes 1 TB per month and charges $15 per TB for the first 10 TB above it (Buzzsprout, read 11 October 2026). Smaller files stretch all of those. Host pricing changes, so read your own plan page. Our comparison of what each host meters is in compressing audio and video for podcasters.

Re-encode the episode at a lower bitrate, in mono if the show is voice only, then check the result against the cap. Smol sets bitrate and sample rate, not output bytes, so the size comes from duration times bitrate and you confirm it.

Two traps. Do not lower the sample rate to save space: in our ffmpeg test at a fixed 48 kbps mono, sample rates from 44.1 kHz down to 16 kHz changed the MP3 size by 0.06%. Keep 44.1 or 48 kHz. And do not expect mono at the same bitrate to shrink the file (see above).

One limitation to know first

Our own test found that Smol drops embedded cover art from audio output on MP3, M4A and FLAC, while text tags came through. That is documented, with the method, in the podcaster guide. If your files carry embedded art, check an output before you publish.

Drop the exported episodes into Smol, choose the output, bitrate and channels, and compress. A whole season goes through in one batch.

1. 1

   Finish editing and loudness first

   Apple targets about -16 LKFS with a true peak no higher than -1 dB FS, applied before encoding. Normalize in your editor, then export a WAV.

2. 2

   Drop the files or the folder into Smol

   Choose Compress, and set the output format: MP3 for reach, M4A or AAC where the host takes it.

3. 3

   Set bitrate and channels from the table above

   Mono at 64 to 96 kbps for voice, stereo at 128 kbps or more for music. Keep the sample rate at 44.1 or 48 kHz.

4. 4

   Compress, then check the sizes and play a minute

   Smol writes a copy with a _compressed suffix; originals are untouched by default. Compare the output size against your host’s cap before uploading.

5. 5

   Save the settings as a preset on this Mac

   Presets live on one Mac. On a second machine, write the settings down and recreate the preset.

Free tools do the job: your editor’s export dialog, Apple Music, or a command line. Smol earns its place only when you batch many files and want settings in a window.

Where Smol is not enough, see the MCP server for agent-driven jobs or the batch audio guide. Related audio work is covered for audiobook narrators and voice-over artists.

Smol is a bitrate and format tool. It is the wrong choice in these cases.

• You need loudness normalization to -16 LKFS. Smol has no loudness feature; use your editor or ffmpeg.
• Your files carry embedded cover art that must survive. Our test showed it dropped on audio output.
• You need to edit, trim or mix. Smol only re-encodes.
• You need a hard byte ceiling. Smol sets bitrate, not output bytes, so you check each result.
• You are not on a Mac. Smol runs on macOS only.

It also has no free trial and purchases are not refundable. The in-browser demo on the homepage is not the app.