Where to generate
the remake.
The pack gives you the frame, the prompt and the measured length. The generator gives you motion. Here is the handoff — eight live tools, official links, no affiliate deals.
The short answer
Unzip the pack. Start with shots.csv and
frames/shot-01-first.jpg. Upload that frame to a generator as the start image,
paste the literal prompt from prompts.md, set the clip length from the CSV,
and generate a take. Then move to shot 02 and stitch the results.
The pack is a shot list, not a movie button. Generating one shot at a time is slower than a single prompt, but it is the only way to keep the cut, the camera and the light from drifting off the reference.
The handoff, in four moves
- Pick the first frame. Open
frames/shot-01-first.jpgfrom the ZIP. This is the start image. If the tool also supports an end frame, use the mid frame from the same shot as a rough target, not as a guarantee. - Paste the literal prompt. In
prompts.md, start with the line labelled literal. Keep the camera, light, palette and duration. Add only the subject and setting you actually want. Save the swap and stylized variants for later re-rolls. - Match the shape. Set the aspect ratio and duration from
shots.csv. Most generators return five to ten seconds even when the reference cut is shorter; generate a little long and trim it in the edit. - Generate two or three takes. Compare the camera direction, the light key and the palette before you compare details. Change one thing per re-roll. Then take the first frame of shot 02 and repeat.
What goes where
| Pack file | Use it for | Put it in |
|---|---|---|
shots.csv | Shot length, aspect ratio, motion and palette in one row. | The generator's duration and aspect settings. |
frames/shot-XX-first.jpg | The composition the new shot starts from. | Image-to-video / start frame. |
prompts.md | The literal, swap and stylized prompt variants. | The prompt box — literal first. |
frames/shot-XX-mid.jpg | A representative still for a look test or an end frame. | Reference / end frame, if the tool has one. |
storyboard.jpg | The whole sequence and the order of the cuts. | Your editing timeline, not the prompt box. |
The shortlist
These are official product pages, checked in November 2026. They are not ranked and they are not the same product: some are better at a controlled camera, some at faces and motion, some at fast look tests. Features, prices and regional access change; open the current page before you buy credits.
Image-to-video with start and end frames, multiple keyframes and motion controls. A practical starting point for people and product shots.
kling.ai (opens in a new tab)Image-to-video plus camera and motion direction inside a production editor. Pick it when you want to adjust the shot, not just re-roll it.
runway.com (opens in a new tab)Google's Veo-based video workspace. Start from a frame, build a scene, and see which model and audio options your account exposes.
flow.google.com (opens in a new tab)Keyframe-driven image-to-video with a cinematic look. Good for a fast look test before committing to a longer take.
lumalabs.ai (opens in a new tab)Image-to-video with expressive motion and longer takes. Useful for character movement and physical action.
hailuoai.video (opens in a new tab)Short-form image-to-video and effects. Fast, playful iterations when exact continuity is not the point.
pika.art (opens in a new tab)CapCut's browser studio: image-to-video and editing in one place. A short hop if you already cut in CapCut.
dreamina.capcut.com (opens in a new tab)Reference-to-video and image-to-video. A second opinion when you need to keep a character or product consistent.
vidu.com (opens in a new tab)Pick by the shot
The right tool depends on the shot in front of you, not on a leaderboard. Start with the job, then test one shot before committing to a batch.
| If the shot needs… | Start with | Why |
|---|---|---|
| A person or product moving from a still | Kling AI, Hailuo AI | Start-frame control and expressive motion; test the current model and plan. |
| A specific camera move | Runway, Luma Dream Machine | These expose camera and keyframe controls; check what your plan makes available. |
| Sound in the same generation | Google Flow | Some Veo models generate native audio; check the model and region your account exposes. |
| A quick, stylized social take | Pika | Effects and fast short-form iterations; continuity is secondary. |
| Generation and the edit in one place | Dreamina | A browser editor around image-to-video; useful if the next step is already a cut. |
| A character or product that has to stay recognisable | Vidu, Kling AI | Reference features help; no model guarantees identity across every shot. |
What to expect — and what not to
A generator will not reproduce the reference frame for frame. It is a new performance of the same brief. The pack gets the measurable decisions right; the model still improvises faces, hands, text and physics.
- Expect drift. Two or three takes per shot is normal. Change one variable at a time.
- Expect to trim. A generator's five-second clip can hold a two-second cut; cut it back in the edit instead of asking for an exact duration it cannot make.
- Expect audio to be separate. Unless the model explicitly generates sound, add music, effects and voice in the edit. The pack's BPM is a guide for where cuts can land.
- Expect continuity work. Use the first frame of the next shot, a reference image or an end frame. No tool keeps a character consistent for free across a whole sequence.
- Do not expect a clone. The pack copies the grammar, not the pixels, the faces or the performances. That is also what keeps the result yours.
If a link no longer opens a creation page, use the service's official site rather than an old tutorial. The category changes faster than any list. For the prompt-writing side, read how to reverse-engineer a video prompt or go back to the workbench and build a pack from your own clip.
Questions
Which generator should I use for a Rea Video pack?
Start with one that accepts an image as the first frame and lets you set a short duration. The shortlist above is a starting point, not a ranking. Pick one, test shot 01, and only then decide whether to pay for more credits.
Can I generate the whole video in one prompt?
No. The pack is a shot list: each shot has its own frame, duration, camera and prompt. Generate the shots separately and stitch them in an editor. Some tools support multi-shot generation, but they still drift from a specific reference.
Do I need to upload the original video?
No. The pack already contains a first and mid frame for every shot. Upload only the frame you choose to animate. Do not upload footage you do not have the rights to use.
Why does the generated clip not match the reference?
Because the model generates a new shot, it does not copy the old one. Start with the literal prompt and the first frame, change one variable at a time, and expect two or three takes. This is a remake, not a forensic match.
Are these affiliate links?
No. The links go to the official creation pages and Rea Video does not take a cut. Availability, pricing and account requirements change; check the official site before you rely on a feature.
Can I use the generated video commercially?
The prompt pack is yours to use. The generated video is governed by the generator plan and terms, plus your rights to the reference material. Check both before publishing.
Start with a clip.
Generate the rest.
The workbench stays free and local. The generator is your call.