Simulon Puts Your AI Characters in the Real World

AI video models forget what your character looks like the moment the camera moves. Simulon 2.0 takes the opposite route: it turns your character into a 3D neural asset, lets you block it in your actual room through your phone camera, and renders the scene from there.

Every creator working with AI video hits the same wall. Your character looks perfect in the first frame — then the camera orbits, and the model quietly decides she is somebody else. Faces drift, outfits change, scale wobbles. The usual fixes are workarounds: turnaround sheets, reference images, prayer.

Simulon takes the opposite route. The app — relaunched as Simulon 2.0 in October 2026 by Beamm Technologies — turns your character into what it calls a neural asset: a 3D subject you place in your actual environment through your phone's camera, then render into the scene with AI handling the lighting, shadows and reflections. Its tagline is "the camera for your imagination," and for once the slogan undersells it slightly.

How it works

You start with a photo, a sketch, or a few words. Simulon builds a subject from it and keeps it in your personal library, ready to reuse across scenes — no 3D or animation experience needed, per the App Store listing. You can also work the other way: photograph a real object, pet or person and "clone reality," turning the photo into a placeable asset, and you can combine two assets into something new.

Directing happens in AR. Place the subject in your surroundings, adjust its size and position — miniature on the table or life-size in the doorway — then move around it to find your angle. From there you have two ways to make video. You can describe the action and let the model perform it, using start and end frames to guide the result. Or, with the feature that landed on October 28, you can act the scene out yourself.

Processing happens in the cloud; the current version (2.1.1) also steadied AR placement and added offline recording, so a take is never lost to a bad connection.

Reality-Aware Rendering

That October 28 feature, Reality-Aware Rendering, is the genuinely new idea here. Co-founder Divesh Naidoo describes it this way: place your assets with your phone and act out the scene; when you stop recording, Simulon renders the take, reading everything in it — what you do, what you say, and the place around you — and your assets respond to all of it. The output, he stresses, is "a new performance of the scene, not a copy of the footage."

In the hands-on video that caught our eye (Theoretically Media's "AI Video Just Stepped Into the Real World!"), you can watch what that means in practice. A real cat wanders across the table mid-scene and the digital character reacts to it. The creator speaks to his character behind a bar and it answers the interruption. A physical flashlight lights the digital cast — and a character picks the torch up and lights the creator's hand in return. None of it is scripted shot-by-shot; the render reads the take and improvises the performance.

Directing, not lottery-picking

Because the character exists in 3D before any video is generated, you get control AI video usually denies you. Multiple characters share one scene at consistent scales, and they interact — in the video, a character works a tiger (one of Simulon's own library subjects) and the two react to each other unprompted. You can break the 180-degree rule that video models routinely mangle, because the geometry of the scene is fixed before generation begins.

For run-and-gun creators that is the appeal: you choose the location, the framing and the staging, and the AI's job shrinks to performance and finish. Simulon outputs also travel well — a scene can go into a video-to-video model to swap the location entirely, which is how the video's creator restyled his footage after the fact.

The honest limits, and what it costs

It is early, and it shows. Placement can be finicky, and generations occasionally go off the rails — a prop appears that was never in the room, a tennis rally turns to slapstick. It is iPhone and iPad only for now (iOS 17 or later); Simulon says Android is coming next.

Pricing matters here, because the free tier is a taster, not the tool: new accounts get 100 one-time starter credits to try subjects from the Simulon library, but creating your own subjects requires a plan. Simulon Plus is $14.99/month and includes commercial use for individuals; Pro is $39.99/month with three times the usage, studio-tier generation and commercial use for businesses. Credit top-ups are available on both plans and never expire.

One caution on the rumor mill: reports of coming character voices, assignable personalities and voice commands come from the review video, not from Simulon — the company has not confirmed them in its own materials. What is confirmed is already plenty.

Where it fits in your backpack

If your content involves recurring characters — a mascot, a persona, an original cast — Simulon is the staging layer your AI video stack was missing: the place where characters stop being references and start being cast. We have added it to the Video Editing section of the directory, where it sits alongside Runway and Pika as a very different answer to the same problem.