A product reveal goes from a closed box (or a cloth, a bag, a dark frame) to the product in full view. On 8frame you give the video model both ends: a first frame with the box closed and a last frame with the product revealed, made so the two frames share the same table, light and camera. You make the last frame from your real photo, build the first frame from it with an edit model such as Flux Edit (7 credits for a roughly 1 megapixel input) or Kontext Max (13), then animate between them on Kling 2.5 Turbo Pro (48 credits for 5 seconds, 1080p) or Kling O3 (3 to 15 seconds). The limit people do not expect: the default Kling node, Kling v3, does not take an end frame on the canvas, and Seedance only does when no reference images are attached.
How first and last frames work on each model, in general, is in AI video first and last frame. This page builds one specific shot: the reveal.
TL;DR
| Order | Node | Credits per run | Connected to it |
|---|---|---|---|
| 1. Product photo | Upload | none | your real product |
| 2. End frame first: the revealed product | Background removal, then Nano Banana Pro | 12; 21 at 1K or 2K | the product photo |
| 3. Start frame from the end frame | Flux Edit, or Kontext Max | 7 for about 1 MP (rises with input size); 13 | the end frame + an instruction |
| 4a. Reveal at 1080p | Kling 2.5 Turbo Pro | 48 for 5 s, 95 for 10 s | start + end frame |
| 4b. Reveal, 3 to 15 s | Kling O3 | 57 for 5 s silent, 76 with sound | start + end frame |
| 5. Sound and end card | Video Editor | no credits | the clip |
The board, node by node
There is no reveal template live on 8frame. The board is a short chain on a blank board: photo, end frame, start frame, video. CAMERA ANGLE is useful if the reveal also needs a change of angle.
Step 1: the product photo
Upload your best photo of the product. The reveal ends on this product, so it is the frame viewers will look at longest.
Step 2: build the end frame first
Counter-intuitively, start at the end. The last frame shows the product, and the product must be right, so make that frame from your real photo: cut it out with Background removal (12 credits), then set the scene on Nano Banana Pro (21 at 1K or 2K) with the cut-out as reference. Include the opened box or the pulled-back cloth in the frame, so the start frame can be made by closing it.
The headphones from the reference photo resting on an open matte black gift box with the lid leaning against it, on a dark walnut table, a single soft spotlight from above, dark background, product centred and fully visible, 16:9
Check the product against the real one: colour, logo, proportions.
Step 3: make the start frame from the end frame
Now edit the end frame into the start frame: same table, same light, same camera, box closed. Ask the edit model to change only the box; then compare the two frames and re-edit until the table, light and camera match, because a shared background is what lets the transition read as one shot.
- Flux Edit (FLUX.2 pro): 7 credits for an input of about 1 megapixel; the price rises with input size, up to 87.
- Kontext Max: 13 credits.
Close the lid of the gift box so the headphones are hidden inside, keep the box, the table, the spotlight and the camera exactly the same
Put the two frames side by side and check: same box position, same shadow direction, same framing. If the box moved, edit again rather than regenerating.
Doing it the other way round (generating a closed box and then trying to "open" it onto your product) asks the edit model to invent the product, which is the one thing that must not be invented.
Step 4: animate between the frames
Connect the start frame and the end frame to a video node that takes both, and prompt the action that connects them:
A hand lifts the lid off the gift box and sets it aside, revealing the headphones inside, the spotlight brightens slightly, camera fixed, single continuous shot
Model choice:
- Kling O3: needs a start image, takes an end frame, any length from 3 to 15 seconds, 720p. 57 credits for 5 seconds silent, 76 with sound (sound is off by default). The choice for a fast 3 second reveal.
- Kling 2.5 Turbo Pro: 48 for 5 seconds, 95 for 10, 1080p. Its files had no audio track in our sample.
- Kling 2.1 Pro (67 for 5 seconds) and Kling O1 Image (73 for 5 seconds) also take an end frame.
- Seedance 2.0 or 2.5: takes an end frame only with a start frame set and no reference images attached.
Step 5: sound and the end card
A reveal is better with a sound: the lid, a soft swell. Kling O3 with sound on generates audio with the clip; an audio track is not proof of good sound, so listen. Or add sound and music on a sound layer in the Video Editor, and put the product name and price on a text layer over the last second.
If you join the reveal to other clips, do it on the Video Editor timeline. The Montage node outputs no audio track.
Three reveal setups, built the same way
Each one starts from the end frame and edits backwards:
- Lid lift. End frame: the product in an open box, lid leaning beside it. Start frame: the same box closed. Prompt the hand that lifts the lid. Good for gifts, cosmetics, electronics.
- Cloth pull. End frame: the product on a plinth with a cloth bunched at its base. Start frame: the cloth draped over the product, edited in with Kontext Max (13). Prompt the cloth sliding off toward the camera. Good for larger objects and launches.
- Light up. End frame: the product lit by a spotlight. Start frame: the same frame almost black, with only a rim of light on the product's edge. Prompt the light rising. This one needs no props at all, and the two frames share everything except brightness, which makes them easy to match.
Whichever you pick, keep the move to one action. A lid lift plus a camera push plus a light change asks the model to resolve three things between two frames.
Which model for which step
| Model | End frame on canvas | Length in this mode | Resolution | Sound | Price |
|---|---|---|---|---|---|
| Kling O3 | yes, with a start image | 3 to 15 s | 720p | off by default | 57 silent, 76 with sound, for 5 s |
| Kling 2.5 Turbo Pro | yes | 5 or 10 s | 1080p | none in our files | 48 for 5 s |
| Kling 2.1 Pro | yes | 5 or 10 s | not listed here | not measured | 67 for 5 s |
| Kling O1 Image | yes | 5 or 10 s | not listed here | not measured | 73 for 5 s |
| Seedance 2.0 | only with no reference images | 4 to 15 s | 480p, 720p | included | 108 for 5 s at 720p |
| Kling v3, 2.6 Pro, 2.1 Master, O1 Reference | no | not usable here | |||
| Wan 3.0 Prime, Pruna p-video | API only | not usable on the canvas |
Where it breaks
- Frames that do not match. If the table, light or box position differs between the two frames, the model invents a cut or a morph. Make the start frame by editing the end frame.
- A generated product. Build the end frame from your real photo. The product is what people look at last.
- The wrong Kling. Kling v3 is the default Kling node and it does not take an end frame on the canvas. Switch the model.
- Seedance with references. Attach references and the end frame is dropped. Choose one or the other.
- Shape set by the frames. Kling nodes follow the start frame's shape. Make both frames in the final shape.
- Morphing instead of revealing. If the box melts into the product, describe the physical action ("a hand lifts the lid") rather than the change.
Platform note
A reveal of a real product in a generated scene is significantly edited media under TikTok's ad policy: use its AIGC label or your own clear disclaimer. Meta labels ads made with its own AI tools and says it will also begin detecting ads made with third-party AI tools.
FAQ
Which 8frame models take a first and last frame? On the canvas: Kling 2.5 Turbo Pro, Kling 2.1 Pro, Kling O1 Image, Kling O3, and Seedance when no reference images are attached. Not Kling v3, Kling 2.6 Pro, Kling 2.1 Master or Kling O1 Reference.
Why should I make the end frame first? Because the end frame shows the product, and the product must come from your real photo. The start frame can then be an edit of it with the box closed, so both frames share the same scene.
How long should a reveal be? Kling O3 takes any whole second from 3 to 15; Kling 2.5 Turbo Pro is 5 or 10 seconds.
Can I make the reveal square? Yes. Kling nodes follow the start frame's shape, so make both frames square.
Why does my box morph into the product? The prompt describes a change rather than an action. Write the physical action that connects the frames.
Sources
- 8frame prices, defaults and limits: the 8frame app, checked 2026-11-09.
- TikTok Ads Policy, Misleading and false content (last updated April 2026): https://ads.tiktok.com/help/article/tiktok-ads-policy-misleading-and-false-content, checked 2026-11-09.
- Meta, "Expanding GenAI Transparency for Meta's Ads Products" (updated 2026-06-01): https://about.fb.com/news/2025/02/gen-ai-transparency-metas-ads-products/, checked 2026-11-09.
Try it without signing up: the 8frame canvas opens on a real board. The canvas is free and unlimited; generation is paid from $19/month.