Reve Text to Image: lock this still before you buy motion (4K, 4cr)

Reve Text to Image is the stills row. Animate after approval. Do not generate the pack inside the video model.

Reve Text to Image is Reve's image model — tight prompt adherence and clean, legible text rendering. It delivers composed, poster-ready frames from dense multi-element prompts. Category: text-to-image. Requires image: no. 4K. 4 credits. Audio: none. Durations: none. The text-to-image tool is the generate. The poster is the test. If you cannot read the still, you will not read the clip.

Type is the reason this row exists

Most stills models are "maybe the words will land." Reve's description is the opposite: clean, legible text rendering, composed, poster-ready, dense multi-element prompts. Use it when the frame is type plus object — a thumbnail, a title card, a pack with a readable lockup.

Prompt adherence here means write the layout: headline, sub, product, background, not a vibe paragraph. Features: Clear, Efficient, Accurate. Styles listed: Realistic, Artistic, Anime. Aspects: 16:9, 3:2, 4:3, 1:1, 3:4, 2:3, 9:16. If the ship is 9:16, generate 9:16. Poster-ready is not "we will crop the 16:9 later and keep the type."

4K ceiling. Qualities: SD, HD, 4K. 4 credits. That is a still you can zoom. Zoom is how you catch the swapped letter before you buy seconds.

Dense prompts, one frame

A dense multi-element prompt is a layout brief. Put the elements in reading order. Do not add a camera move; there is no duration field. Do not add a voiceover; there is no audio.

If the still would fail as a poster, it will fail as a 10-second product turn. Motion does not fix kerning. Motion makes bad type dance.

No image required. This is a generate, not an edit. If you already have the poster and you only need the headline swapped, use an edit row. Reve will happily sample a cousin poster that is not the one you cleared.

Then — and only then — animate

Sequence:

1. Generate the 4K still on Reve. 4 credits.
2. Read every word at phone width.
3. Lock the frame.
4. Hand it to image-to-video.

Best AI image generator ranks the category if you needed a different stills look. This catalog page is the spec: 4K, 4 credits, text-to-image, legible type, no motion.

Do not generate the pack inside the video model. The video model will invent the letters. Reve's job is to lock the letters first.

FAQ

Is Reve Text to Image a video model?

No. Text-to-image, 4K, 4 credits, no durations, no audio. Poster-ready stills with tight prompt adherence and legible type. Animate after you can read the frame.

Why lock the still if I already know the motion?

Because the motion row will not typeset your headline. If the pack is words plus object, the words have to survive as a JPEG. Reve is the 4-credit test of that.

Does it need a reference image?

No. Requires image is false. You write a dense prompt. Editing a locked poster is a different category.

What should I check before I buy motion?

Read the type. Check the product. Check the crop at the ship ratio. If the 4K still would not post, do not spend a video credit making it move.