Suno Sounds V5.5: a voice file, not a talking generate (2cr)

Suno Sounds V5.5 writes audio. If we see the mouth, pick a talking or lipsync row instead of laying this on a closed mouth.

Suno Sounds V5.5 is the latest sound generation model with improved quality, looping, tempo, key controls, and lyrics subtitle capture. Content type: audio. Category: text-to-audio. Audio: true. Credits: 2. No resolution. No durations listed. Requires image: false. The assignment door is AI text to speech; the neighboring tool that matches a bed is AI music generator. Neither is a face.

This row does not move lips

You get a file. Looping, tempo, key, lyrics subtitle capture — those are mix controls, not a mouth. Features: sound_generation, looping, tempo_control, key_control, lyrics_capture.

If the cut shows a person speaking, this 2-credit file on a closed mouth is a dub you will hate. Pick a talking or lipsync row so the mouth is a job, not a coincidence. If the cut is B-roll, a title card, a product turn, then yes: generate the bed here and lay it under the picture.

Do not prompt "a woman talking to camera" on Suno Sounds and expect a take. You will get sound about a woman. The picture is someone else's row.

Two credits for the bed, not the episode

2 credits. Flat billing. That is why people generate six versions of a loop. Fine — if you are actually listening. Tempo and key are in the description so you can match the cut, not so you can ignore them and stretch a 120bpm loop over a 70bpm voiceover.

Lyrics subtitle capture is for when the sound has words you might caption. It is not burned-in captions on a video. If you need on-screen type, that is an editor. If the lyrics are the whole piece and there is a mouth in frame, you are back to the claim: talking row, not this file on a closed mouth.

No aspect ratios. No 4K. No 5-second tick. The catalog is telling you this is not a clip.

Picture first or picture never — never mouth-as-afterthought

Sequence that ships:

1. Lock the picture (still, clip, or silent video row).
2. If we see a mouth that must match, leave this page.
3. If we do not, generate the 2-credit bed on V5.5. Set tempo and key. Loop if the cut loops.
4. Mix under, do not replace, unless you meant to strip the live sound.

If you needed a spoken line instead of a bed, that is the text-to-speech tool on a speech engine — still not a face. This page is V5.5: text-to-audio, 2 credits, looping and key, a file, not a talking generate.

FAQ

Can I use Suno Sounds V5.5 to make a talking-head video?

No. Content type is audio. Category is text-to-audio. 2 credits, no resolution, no clip durations. You get a sound file. Mouths need a talking or lipsync row.

What is 2 credits buying?

A sound generate with improved quality, looping, tempo, key controls, and lyrics subtitle capture. Not a 4K twin. Not a 10-second scene.

Should I lay this under a closed mouth?

Only if the mouth is not supposed to be speaking. If we see speech, pick lipsync. Laying a vocal on a still jaw is the mistake the claim names.

Does it need an image or a video upload?

No. Requires image is false. Text in, audio out. The picture, if any, is a different model.