Simulon Puts AI Characters in Your Living Room With AR
Simulon's mobile app places 3D AI characters into your real environment using reality-aware rendering. Here's how it works and what it costs.
What is Simulon?
Simulon is a mobile app that turns a character reference image into a 3D asset you can place, scale, and rotate inside your actual physical environment using your phone’s camera, then generate video of that character acting inside the real scene. The company first showed up in 2023 with early AI character tech and has now returned with a mobile AR workflow that blends a 3D character pipeline with camera-based scene placement, something it calls reality aware rendering.
TL;DR
- Simulon lets you upload a character reference image, converts it into a 3D asset, and lets you place it directly into a real-world camera feed via your phone.
- The placement step is handled through touch controls (two-finger rotation, scaling) and can feel finicky at first, though it becomes manageable with a bit of practice.
- Generations are not real time. You prompt for an action or camera move and the system renders video afterward rather than animating live.
- The standout feature is reality aware rendering, where characters visibly respond to real physical input, like a real flashlight illuminating a digital character or a character reacting to a spoken voice command.
- Multiple characters can appear in a single scene together, including unscripted interactions between them that weren’t explicitly prompted.
- Simulon is free to try, with paid subscription tiers at $14 and $39, and is currently iOS only while Android approval is pending.
- Voices and assignable character personalities are planned additions, which would let two AI characters placed in the same room interact with each other autonomously.
How does Simulon’s AR character placement work?
The workflow starts with a single reference image of a character, anything from a hand-drawn design to a previously generated AI image. Simulon converts that reference into a 3D model. From there, you open your phone’s camera and place the asset into your real surroundings, adjusting its scale and rotation so it sits convincingly in the room, on a table, in a chair, wherever you want it.
Once the character is placed, you prompt for an action, a line of movement, or a camera move, and Simulon generates a video based on that placement. Because the 3D model is locked into the scene’s real geometry, your own camera movement becomes part of the shot. Panning across a room or walking toward the character carries over into the final output, since the system is tracking the actual physical space rather than hallucinating a new one, which is the approach taken by AI world models.
That distinction matters. Reference-based AI video tools can already use a character image to generate motion in a generic generated environment. What Simulon is doing differently is anchoring the character to a real, specific location captured through your camera, letting you block shots the way you would on an actual film set.
What is reality aware rendering?
Reality aware rendering is Simulon’s term for making digital characters respond to real-world stimuli rather than just following a text prompt. In practice, this has shown up in a few ways: a character reacting to a real flashlight being pointed at it, with the light appearing to physically illuminate the digital figure, and characters responding to spoken voice commands directed at them during filming.
This is a meaningfully different interaction model than typical AI video generation, where everything in the frame is driven purely by a prompt. Here, the character is meant to behave as if it’s actually present and aware of what’s happening around it in the real environment, not just inserted into a static background.
It’s also unpredictable in interesting ways. When multiple characters are placed in the same scene, they’ve been observed interacting with each other without being explicitly told to, suggesting the system is making some contextual judgment calls about how characters in proximity should behave.
Can you put multiple AI characters in one scene?
Yes. Simulon supports placing more than one character into the same environment at once. Each character still starts from its own reference image and gets its own 3D asset, but they can be scaled and positioned together in the same real-world shot, then prompted to act within that shared scene.
Results are inconsistent depending on the complexity of the action. Simpler interactions tend to hold together well, with both characters staying scale-appropriate and spatially consistent across the shot. More demanding physical actions, like two characters engaged in rapid movement, can come out looking rough, with less stable tracking. That’s consistent with where AI video generation generally still struggles: fast, complex physical motion involving multiple subjects.
One technical note worth flagging for anyone building with this kind of tool: Simulon’s camera-anchored approach appears able to break the 180-degree rule, a camera convention AI video models typically have a hard time respecting, since the system is tracking a real continuous camera move rather than inferring spatial continuity on its own.
Is Simulon worth trying, and what does it cost?
Simulon is free to try, which makes it low-risk for anyone curious about AR character placement or the broader 3D-to-AI-video pipeline. Paid subscription tiers are available at $14 and $39 for users who want more generation capacity or features beyond the free tier. The app is currently iOS only, with Android availability pending platform approval.
For builders already working in the 3D-to-video pipeline, generating 3D models in tools like Trippo 3D, Hunyuan 3D, or Meshi and bringing them into something like Blender, Simulon offers a faster, more casual path to the same basic idea: get a consistent 3D character into a real or rendered scene without manually rebuilding the whole pipeline. It won’t replace a full 3D production workflow, but as a fast way to block shots, test character placement, or generate quick social content, the free tier is an easy way to see if the approach fits your workflow.
What are the current limitations?
Generation is not real time. You place a character, issue a prompt or camera direction, and wait for the system to render the result, so this isn’t live AR performance capture. The placement controls (rotating and scaling assets with touch gestures) take some getting used to. And output quality is inconsistent: most of the time a character’s position, scale, and presence in the scene hold up convincingly, but occasionally the model introduces elements that don’t exist in the real environment, or physical interactions (like two characters moving quickly together) come out visually unstable.
Voice and personality features, which would let a character have an assigned personality and respond conversationally, are planned but not yet fully available. Once live, that would extend reality aware rendering further, allowing two AI characters placed in the same physical space to interact with each other based on their own defined personalities rather than a single prompt.
Frequently Asked Questions
What platforms is Simulon available on?
Simulon is currently available on iOS. Android support is planned and pending approval on that platform.
Does Simulon generate video in real time?
No. You place a character in your environment and prompt an action or camera move, and the video is generated afterward rather than rendered live as you interact with the scene.
How is Simulon different from standard AI video generation tools?
Most AI video tools use a reference image to generate motion inside a generated or hallucinated environment. Simulon anchors a 3D character to your actual real-world camera feed, so your real camera movement and surroundings carry directly into the generated output.
Can Simulon characters react to things in the real world?
Yes, through what Simulon calls reality aware rendering. Characters have been shown responding to real physical light sources and to spoken voice commands, rather than only following written prompts.
How much does Simulon cost?
Simulon offers a free tier to start, with paid subscription plans priced at $14 and $39 for expanded access.