Turn stills into a slideshow, not T2V

Sequence the stills. Then convert. A slideshow is a stack of frames with a hold time, optional overlay, optional music, then a render to MP4. It is not a video model. Turn a slideshow into a video (convert_slideshow_to_video) needs a slideshow_id. Sending “a six-slide carousel about morning habits, cinematic, 4K, native audio” to text-to-video is a new film that happens to be about a list. You will not get six cards you can reorder.

This page is the stills-to-board-to-MP4 path. It is not a slideshow is not a video model restacked, and it is not a T2V prompt guide.

Stills first, in an object you can edit

The object has:

• N stills, in an order you can edit.
• A hold. duration_per_image on convert is one number for the set, not per-slide storytelling.
• Optional overlay captions (copy stamped on top) or type baked into the image by an image model.
• Optional voiceover_url / music_url that already exist. Convert does not write a track.
• An aspect. Set the ratio on the stills, not later in a video prompt.

What it does not have:

• A duration ladder like 4s / 6s / 8s on a generate row. Those seconds are a clip. A slideshow’s seconds are N × hold.
• Native audio from a video model. A bed is a file you attach.
• Motion control, lipsync, first-last interpolation. Those are other slugs.

Three honest sources for the stills:

1. Photos you already have. Product shots, archive scans, screenshots of the UI. Turn photos into a slideshow arranges your images.
2. Images you generate as stills. One image row per card, approved for type and SKU, then added to the board. Priced per image, shown before you confirm. That meter is image credits, not per-second video.
3. A planned board from a prompt. Generate a slideshow from a prompt plans stills, generates or curates them, stamps overlay if you asked. You still get a slideshow_id, not a Veo take.

Fifteen of nineteen public boards on /slideshows are six frames at 36 credits. That catalog is a scene graph — slide count, cost, prompt, aspect — not a vibe. Recreate a graph. Do not ask Kling to be a carousel.

Approve each still as if it will be paused. On mute, a 2.5-second hold is a poster. If the SKU is unreadable on the still, convert will not sharpen it. If slide three is the wrong product, fix slide three before you render. That is a different page; the point here is that you can fix it because you still have a board.

Convert is a render, not a generate

Turn a slideshow into a video:

1. get_slideshow — confirm the live order and the slideshow_id. Slide N is the Nth image after sorting by order_index ascending.
2. Set duration_per_image and transition. Short holds suit a hook-led reel. Longer ones suit an explainer. One value for the whole set. Per-item timing is compose_video_with_overlay, not this tool.
3. Attach music_url and/or voiceover_url if you have files. Generate them first (generate_music, generate_speech) if you do not. Convert will not write a chorus because the prompt said lo-fi.
4. Set use_edited_images if you stamped overlay. Forget the flag and the MP4 is the bare originals. A reel that works on mute needs the stamped versions.
5. Set aspect_ratio and output_resolution. The same board can go out as 9:16 and 1:1. That is two renders, not two T2V jobs.

The slide images were already paid for when the slideshow was generated. Convert is a render pass. Music and VO are billed on their own tools. Ask for estimate_cost before a long set.

AI slideshow maker is the stills door. AI video generator is the clip door. Mixing them is how a list becomes a scene you cannot reorder.

If you needed a six-second push-in on one product still, that is image-to-video, not a 3-second crossfade across eight cards. Pick the job that matches the motion you actually want. A slideshow is not a slow Ken Burns on a single hero unless you only put one still on the board — and even then, convert is a hold plus a transition, not a camera generate.

Why T2V is the wrong first click

A video model invents a continuous take. It will:

• Blend six ideas into one kitchen.
• Refuse to let you swap card three without regenerating the clip.
• Burn any “captions” as pixels in the scene, uneditable.
• Charge a video sticker (Veo 3.1 is 40 credits for 4/6/8s at 4K with native audio; Seedance 2.0 is 69 credits) for a job that was image credits plus a render.

YouTube’s inauthentic-content buckets (Mashable, 21 July 2026) include generic/repetitive slideshows and scrolling text that rehash. AI itself stays monetizable when original. A board of six approved stills with a real offer is a product. A T2V “listicle cinematic” that is the same as last week’s with the names swapped is the policy risk. The format is not the violation. The sameness is. Convert does not exempt you. It also does not create the sameness by itself.

Do not prompt a video row to “be a carousel.” There is no carousel mode on Veo. There is a slideshow object, then convert.

After the MP4 exists, captions on the video are a second burn if you still need word-timed type on top of overlay. Overlay on the stills is the mute-first copy. Video captions are the speech track, if you attached one. Do not double the same sentence in two layers unless you meant to.

Lock the board, then ship two shapes if you must

If this week’s destinations include a carousel and a reel, you already have the stills. Post the board as a carousel. Convert a 9:16 render for the reel. That is one set of image credits, two deliveries. A T2V 9:16 and a T2V 1:1 are two video bills and two different pictures.

If you need a seat to build the board and run convert, sign up. Bring the stills. Do not bring only a paragraph about six slides.

FAQ

Can convert invent slides from a sentence?

No. It renders a slideshow you already have. convert_slideshow_to_video needs a slideshow_id. Plan or upload stills first.

Will my overlay text appear in the video?

Only if you set use_edited_images. Without it the render uses the unstamped originals.

When is T2V the right tool instead?

When you want a continuous shot — a camera move, a performance, native-audio dialogue — not a stack of cards. A list you must reorder is a slideshow.