Introduction

Hello, real-world directors! On Oct 28, Simulon (from Beamm Technologies) switched on Reality-Aware Rendering in its iPhone app. It's a closed, app-only video generation feature: you place a 3D "neural asset" in your surroundings with AR, film yourself acting out the scene, and when you stop recording Simulon regenerates the whole take so the creature responds to your actions, your voice, and the space around it.

The launch post from Divesh Naidoo passed 900K views in under a day, and creators are already filling X with dragons in shops, kaiju on city streets, and aliens on hikes. You can run it today on iPhone, with Android listed as "coming soon."

What shipped

Reality-Aware Rendering in Simulon 2.0

The idea is simple and a little magical: instead of typing a prompt and hoping, you show it, say it, and perform it.

  • Place: drop a neural asset into your space with your phone. Its 3D placement sets the starting position and scale.
  • Perform: hit record. The asset turns into a "ghost" while you act, gesture, and give spoken instructions.
  • Render: when you stop, Simulon reads everything in the recording (what you do, what you say, and the place around you) and generates the asset's behavior to match.
  • Result: the final is regenerated from your recording. Simulon describes it as "a new performance of the scene, not a copy of the footage."

In an early breakdown, Naidoo placed a raptor next to water, dropped a stone in, and the raptor reacted to the splash. Simulon's own launch short was made "in an afternoon," with "zero prompting," using only real footage, 3D neural assets, and Reality-Aware Rendering.

Where the assets come from

Simulon 2.0's other tools feed the same pipeline:

  • Clone reality: turn real photos into 3D neural assets.
  • From sketch to real: turn a drawing into a placeable asset.
  • Combine assets: merge two assets into something new.
  • Guide the video: edit the scene, describe an action, or use start and end frames.

Pricing and where to run it

Simulon is a closed, subscription iPhone app (no open weights, no public API). From the Simulon site:

  • Free: 100 one-time starter credits, using subjects from the Simulon Library. No subscription needed.
  • Simulon Plus: $14.99/month. Create your own subjects from words or images, turn scenes into realistic video, commercial use for individuals.
  • Simulon Pro: $39.99/month. 3× Plus usage, studio-tier video generation, up to 15% off top-ups, commercial use for businesses.
  • Top-ups: 100 credits for $4.99, 260 for $12.99, 500 for $24.99 (they never expire).

Android is "coming soon," per both the site and the launch post.

A first scene to try

Here's a starter you can act out in five minutes:

Place a small friendly dragon on your kitchen counter next to a bowl of fruit. Start recording, slowly slide one grape toward it, say "go on, try it," then pull the grape back at the last second and laugh.

Keep your gestures clear and big, give one spoken cue at a time, and hold the phone steady so the scene reading has a clean plate to work from.

Why it matters

Most AI video still starts with a text box. Reality-Aware Rendering flips the order: your timing, your blocking, and your actual room become the prompt. For indie filmmakers, VFX tinkerers, and short-form creators, that's a big step toward directing AI characters like actors instead of rolling dice on prompts.

Original Source

Launch thread by Divesh Naidoo on

, plus pricing on simulon.com.

Conclusion

Simulon just made "film it yourself, let the AI perform" a real button on your phone. Grab the free starter credits, put something impossible on your counter, and see how it reacts to you.

—Aurelia ♡