Suno custom models on Pro vs Sume Music: no fine-tune

Suno's pricing page lists custom model creation as a Pro plan feature. Sume has no way to train or fine-tune a music model: the Music Router takes a prompt and optionally one image, so a consistent brand sound comes from a saved brief and a reference still.

What the vendors list

Suno names custom model creation under Pro and Premier (read 2026-10-02). ElevenLabs documents Music Finetunes, where you upload non-copyrighted tracks and get a personalized model in about 5 to 10 minutes (read 2026-10-02). Sume documents neither.

A repeatable house sound without training

The Music 1.0 docs suggest a seven-axis brief: emotion, genre lineage, tempo as a number, key and mode, two to four instruments with texture, an arc with one named moment, and era or production. Save that text as your house template and change only the arc line per project.

For a multi-scene project the docs say to preserve continuity when one score is requested, and to pass the accepted scene still as image_url. They also say to vary genre family, tempo by at least 12 BPM and lead instrument when scenes should contrast.

Limits to know

There is no seed, so the same brief gives different tracks each time. Treat the template as steering, not as a guarantee. Music Router also rejects a non-empty negative_prompt with a 400, so write exclusions as positive text such as no vocals.

If your brand needs the model itself to learn a catalog, that is a training feature Sume does not offer. Use a vendor that does, and bring the finished audio into a Sume Timeline render as the soundtrack.