ElevenLabs Music output_format list vs the file Sume returns
ElevenLabs Compose lists MP3, PCM, mu-law, A-law and Opus formats. Sume's music request has no format field; get wav through Timeline audio.

ElevenLabs Compose accepts an output_format with MP3, PCM, mu-law, A-law and Opus variants. Sume's music request has no format field: the completed job returns an audio artifact, typically audio/mpeg, and you convert to wav with a Timeline audio job.
What ElevenLabs lists
The Compose reference lists auto as the default plus MP3 at several sample rates and bitrates, PCM from 8000 to 48000 Hz, ulaw_8000, alaw_8000 and Opus at 48 kHz (read 2026-10-02). The capabilities page says output is MP3 (44.1 kHz, 128 to 192 kbps) and WAV (read 2026-10-02).
| Source | What it says about format |
|---|---|
| ElevenLabs Compose reference | output_format: auto default; mp3, pcm, ulaw, alaw and opus variants |
| Gemini API Lyria 3.5 | MP3 by default, WAV available for Lyria 3.5, 44.1 kHz stereo |
| Sume Music Router | No format field; audio artifact on media.sume.com, typically audio/mpeg |
Getting wav on Sume
Timeline audio takes Sume-hosted audio and returns a durable file. Its output.format is wav by default (pcm_s16le) or mp3. A concat with one part is a valid input, since parts takes 1 to 20 items, so it works as a conversion step.
Note the docs warn that mp3 output re-adds priming padding at every edge, so keep wav for anything you will cut again.
Telephony and low-bitrate targets
Mu-law, A-law and 8 kHz PCM matter for phone systems. Sume's music docs list none of those, so for hold music you generate the track and convert it in your own pipeline. The Sume Timeline docs do not cover mu-law either.
For web delivery, an mp3 file from a Music job usually needs no conversion. Do not assume a sample rate: inspect the artifact before you hand it to a platform with strict audio specs.
Sources
Related posts
More in Developers
- ElevenLabs Music v2.5 became the default: which model made a track?
ElevenLabs made Music v2.5 the default on 2026-09-11. When a default changes, old and new tracks differ. How to record the engine per track, on Sume too.
- elevenlabs say CLI: text to speech from a terminal, and Sume's route
ElevenLabs' CLI got a say command (Eleven v3 default) on 2026-09-07. Sume's CLI has no audio generation commands yet; the TTS job is a plain HTTP call.
- ElevenLabs v4 10,000-character limit vs Sume TTS 20,000 transcript
Higgsfield's changelog lists 10,000 characters per ElevenLabs v4 generation; Sume TTS 1.0 takes 20,000. How to split a long script.
- Export video model prices weekly: a catalog-to-CSV script
Prices and limits for AI video models change. A short script reads Sume's video catalog into a CSV so you can date each row and track changes.
Written by Sume