Suno v6 audio and video input vs Sume: text and one image

Of the four input types Suno v6 names, Sume's music models accept two: a text prompt and one optional image. Audio and video are not inputs to POST /v1/music-1.0/generate or the Music Router. The request has no field for them.

Suno's list is from its v6 announcement. Sume's facts are from Music 1.0 and Music Router, read 2026-09-30.

What inputs does Suno v6 list?

The announcement says you can "Create with text, audio, images and video", and also lists building a mashup from multiple sources and sampling or isolating audio to make a new beat. It does not give limits for those inputs.

Which of those can I send to Sume?

Two of the four. The docs say a request is a text prompt plus optional image conditioning, and the image must be public HTTPS. A music request carries nothing else that could hold a sound or a clip.

Can I use a video's sound some other way?

Sume can pull the audio track out of a hosted video with audio detach, and join or split Sume-hosted audio with Timeline audio. Those produce audio files; they do not feed a music request, which has no audio field.

What should I do for a picture-led track?

Describe the mood, tempo and instruments in prompt and pass the still as image_url. For the image-input side across engines, see Lyria 3.5 reference images vs Sume's image_url.