Mubert Render 25-minute tracks and licence limits vs Sume Music

Mubert Render can make tracks up to 25 minutes on paid plans, while a Sume Music generation is a song-length track of up to a few minutes that you lengthen by looping inside a timeline. Mubert also forbids publishing its tracks on streaming platforms or registering them in Content ID, which matters if you plan to release the music rather than lay it under video.

Mubert facts are from its Render pricing page, read on 2026-10-02. Sume facts are from the Music 1.0 and Timeline 1.0 docs.

What does Mubert Render list?

Mubert says it generates royalty-free music from text prompts or images, with genre, mood, BPM and instrument controls. All paid plans support tracks up to 25 minutes; the free Ambassador plan offers only fixed-duration tracks.

Its licence limits are stated plainly: tracks cannot be published on music streaming platforms (Spotify, Apple Music, Deezer, YouTube Music and similar) and cannot be registered in any Content ID system. Commercial use requires the Pro or Business plan; Creator is restricted to non-commercial projects.

How long is a Sume music track?

Music 1.0 runs on Google Lyria 3.5 and makes full-length structured songs up to a few minutes. There is no duration or duration_seconds field (both are rejected), so you steer length in the prompt, for example "a 2-minute track" or section markers such as [0:00-0:30] Intro: .... The Music Router resolves to the same engine today.

The price is a fixed $0.125 per accepted generation, whatever the prompt length or whether an image is attached. There is no 500-track monthly bundle to compare against: Mubert's Creator plan works out to about 2.3 cents a track if you use all 500, and Business to about 15 cents, while Sume is always 12.5 cents. At small volumes the difference is cents; at 500 tracks a month it is about $62.50 versus $11.69.

How do I get a 10-minute bed on Sume?

Generate one good track and let the render loop it. In Timeline 1.0 the soundtrack field takes url, gain_db, loop, fade_out_seconds up to 10 and duck_db, and the spine can be up to 1,800 seconds long. Looping repeats the same track, so listeners may hear the seam on a long piece. For variety, generate two or three tracks with different tempo or lead instrument and join them with timeline audio concat.

curl -X POST https://api.sume.com/v1/timeline-1.0/audio \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: long-bed-001" \
  -d '{
    "operation": "concat",
    "parts": [
      { "url": "https://media.sume.com/artifacts/artf_demo/bed-a.mp3" },
      { "url": "https://media.sume.com/artifacts/artf_demo/bed-b.mp3" },
      { "url": "https://media.sume.com/artifacts/artf_demo/bed-c.mp3" }
    ],
    "output": { "format": "wav" }
  }'

What about licensing and releasing the music?

Mubert's page spells out what you may not do with its tracks. The Sume Music pages we read do not give a licence summary, so do not assume anything either way: if you plan to distribute a track as a song, check Sume's terms and the rules of the platform you are releasing on. Sume's docs do state that Music 1.0 and the router call the same engine, Google Lyria 3.5, so provider terms for that engine are also worth reading.

If you only lay music under video, the question is smaller. Either product can make a bed; the practical difference is the track length per call and the price model.

Which should I choose?

Choose Mubert when you need single tracks of ten to 25 minutes and you accept its streaming restriction. Choose Sume when music is one step in a video build, you want every call on one wallet next to voice and render jobs, and a looped or concatenated bed is fine.

Before committing, test the one thing that differs most: how a long piece holds up. Generate a Sume track from a brief that names a tempo, a key and a moment where the arrangement changes, loop it under a few minutes of your real video, and listen for the seam. Do the same with a Mubert track of the same mood and length. The cost of that test is $0.125 on Sume and one of your monthly Mubert tracks, so it is cheaper than a wrong assumption about either product.