AI Squish, Inflate, Melt and Explode Video Effects

VdoBloom’s viral effect tabs turn one uploaded photo into a physics gag: squish squeezes an object like memory foam and springs it back, inflate blows it up like a balloon, melt sags it into a glossy puddle, and explode bursts it into confetti. Each effect starts from a single image and returns a downloadable MP4 in the duration, resolution and aspect ratio you choose. They are built for objects, not people.

Updated October 2026.

What each of the four physics effects does

Squish is the flagship. Two hands enter from the sides and squeeze the object like soft memory foam; it bulges between the fingers, flattens, then springs back to its exact original shape when the hands release and leave the frame. The object keeps its colours, texture and detail throughout, the camera stays locked, the lighting stays constant, and only one action happens.

Inflate makes the object blow up like a balloon. Melt softens it so it sags and drips into a glossy puddle of its own colours, with no smoke or fire, one action and a locked camera. Explode, labelled Confetti Explode, bursts the object into confetti. All four start from one uploaded photo, so the input is the same across every effect.

One detail separates these effects from a generic animation prompt: each is built to describe a single clean action rather than a chain of events. Squish does one thing — squeeze, hold the deformation, release — and the return is the payoff. Melt does one thing: soften, sag, pool. Because the camera is locked and the lighting stays constant, the viewer reads the change in the object and nothing else, which is exactly what lets the clip loop cleanly on a short-form feed.

Step by step: from one photo to a finished clip

1. Open the effect tab for the effect you want — inflate, melt or explode, or any other from the viral effects index.
2. Upload a single photo of the object. One image is the whole input.
3. Choose the duration, resolution and aspect ratio before generating. The credit cost reflects those choices.
4. Generate the clip and watch for the key beat: the squish must release and return the object to its exact original shape.
5. Reuse the same photo on the other three effects. One still works across all four, which is the cheap way to build a consistent set of clips for a single product.

The four effects compared

Effect	What happens	Ends back in original shape?
Squish	Two hands squeeze the object like memory foam; it bulges, flattens, then springs back	Yes, to its exact original shape
Inflate	The object blows up like a balloon	No
Melt	The object softens, sags and drips into a glossy puddle of its own colours	No
Explode	The object bursts into confetti	No

Making a series from one photo

The same uploaded photo works across all four effects, and that reuse is the single biggest efficiency in the workflow. Shoot or source one clean product image, then run it through squish, inflate, melt and explode in turn. You get four distinct clips that share framing, colour and product, which is what makes a short campaign look intentional rather than assembled from unrelated footage. It also removes the usual risk of an effect clip — that the object looks different from the rest of the set — because the object literally is the same image every time. Build the set in one sitting and the products, light and background match by construction.

Composition rules that make the shot work

• Leave space at the left and right of the frame for the hands to enter.
• Shoot the object at roughly two thirds of frame width.
• Use a plain, unbroken surface so nothing competes with the object.
• Keep a soft contact shadow so the hands and the object sit on the same surface.

These come from the tool’s own guidance and they matter more than they look. A frame where the object fills the whole width leaves the hands nowhere to come from, and a busy background makes the physics read as a glitch rather than a gag.

Which photos work and which fail

Objects with a clear silhouette and a single colour or texture work best: a plush toy, a sneaker, a bottle, a packaged snack, a phone. The effects need the model to read the object as one mass, so a clean edge and even lighting help. Photos that fail tend to share a few traits: several similar objects the model cannot tell apart, an object cropped by the frame edge, hard directional shadows that flatten detail, or a cluttered background. Faces and people are the wrong input entirely — these effect tabs are built for objects, not people, and the results on human subjects are unpredictable.

If you are shooting the photo yourself, light it flat and bright, and avoid strong coloured gel light, which can tint the deformation and make the object look like a different product. A soft overhead source over a large white surface gives the model the fewest surprises to resolve. If you are using existing packshots, pick one taken straight on rather than at an angle, so the squeeze and the drip happen across the full width of the object instead of along a foreshortened edge.

What it costs

Every effect draws credits from the same balance as the rest of the platform. Signing up gives you 10 free credits with no card needed. Lite is $15 per month for 300 credits, or $10 per month billed yearly, and one-time credit packs start at $10.39 for 200 credits and never expire, with India plans starting at Rs.199 per month. Because each effect runs at the duration and resolution you pick, a longer, higher-resolution clip costs more credits than a short, low-resolution test — the effect tab itself does not carry a separate per-second price. The exact cost is shown before you generate, and paid output is watermark-free with commercial rights.

A sensible budget habit is to preview at the shortest duration and lowest resolution you can still judge the motion at, then re-run only the winner at the final format, since only the final run needs the higher cost. The free 10 credits cover the first experiments, and because a failed generation is refunded automatically rather than charged, a rejected take does not eat into that balance.

Which model to pick

There is no model switch to agonise over here: the tabs differ by physics, not by engine, and you choose duration, resolution and aspect ratio before generating. The practical decision is which effect fits the product. Squish is the one with a return beat, so it suits a reveal where the object must survive intact. Inflate, melt and explode are one-way gags that end with the object changed. For a social cut, reuse the same photo across all four so a short series looks like it came from one shoot.

Frequently asked questions

Do all four effects use the same photo?

Yes. A single uploaded photo can be reused across squish, inflate, melt and explode, which is the cheap way to build a consistent set of clips for one product.

Do these effects work on people?

No. The physics effects are built for objects, not people. Faces and human subjects give unpredictable results, so use a clean, clearly shaped object instead.

Does the squished object return to normal?

In the squish effect, yes. The object bulges and flattens, then springs back to its exact original shape when the hands release and leave frame. Inflate, melt and explode do not restore the object.

What format is the output?

A downloadable MP4, in the duration, resolution and aspect ratio you selected before generating.

How much do the effects cost?

They draw on the same credit balance as every other tool, and the exact cost is shown before you generate. Longer, higher-resolution clips cost more credits than short, low-resolution tests.

Pick one photo, then run it through each effect to see which gag fits the product. Start with the squish effect and reuse the same image across the rest.