> ## Documentation Index
> Fetch the complete documentation index at: https://docs.reactor.inc/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> To build and serve your own model, start at /deploy/development/quickstart and /deploy/development/overview. Deploying is the default path: reactor init scaffolds a workspace, reactor auth login authenticates, and reactor model deploy registers the model, publishes the release with the weights/ folder, and activates it on Reactor's GPUs, in one command from that workspace. Docker must be running, because the publish step builds the image locally. Bump model.version in reactor.yaml before redeploying a change, because a release that already has an image is reactivated as it is. Deployment access is granted per account, so contact team@reactor.inc if a deploy is refused. Every key in reactor.yaml is documented at /deploy/platform/reactor-yaml. Model code imports reactor_runtime; Python client code imports reactor_sdk. The runtime overview explains the model interface. Running the model on your own machine with reactor run is optional and needs a GPU you attach with --gpus; /deploy/development/local-testing covers that loop and pairs a complete brightness model with a Python client test in a separate brightness-test workspace.
> Reactor hosts multiple models, each with its own connect slug (modelName) and command/event schema. The video model catalog — slug, typed SDK package, and links to its schema — is at /model-api-reference/overview. Robotics policy documentation starts at /robotics/overview; X-WAM observations, actions, and client integration are under /robotics/xwam/; Cosmos3 Nano Policy DROID is under /robotics/cosmos/nano-policy-droid/. Some models expose one slug per experience (e.g. HappyOyster); always take the slug from the model's own pages, never guess it.
> Fastest path to a working app: `npx create-reactor-app my-app --model=<slug>` scaffolds a complete app with secure auth wired up. Typed TypeScript SDKs are published as @reactor-models/<model>; Python uses the base reactor-sdk package.
> Auth: exchange an API key (rk_...) for a JWT via POST https://api.reactor.inc/tokens from your server. Never put the API key in client-side code.
> Append .md to any docs URL for clean Markdown. Search these docs via the MCP server at https://docs.reactor.inc/mcp.

# Cosmos3 Nano Sim Bimanual

> Move virtual robot arms and see a learned manipulation scene respond.

Cosmos3 Nano Sim Bimanual generates video in response to arm movements and gripper controls. Start
with an image of the supported robot and scene, describe the task, and steer its arms while watching
the generated observations.

Use it to build an interactive manipulation playground: reach across a table, open or close a
gripper, and reset the scene to explore another movement.

## What you provide and receive

| You provide | You receive |
| - | - |
| A starting image, task text, and calibrated arm poses and grips | A continuous stream of generated RGB observations |

## Supported embodiments

| Embodiment | Controls | Get started |
| - | - | - |
| **AgiBot** | Calibrated arm poses and gripper values | [AgiBot overview](/robotics/cosmos/nano-sim-bimanual/agibot/overview) |

The hosted checkpoint supports AgiBot. Other robot embodiments are not supported by supplying a
different starting image or substituting their actions. Choose the embodiment guide for input
conventions and integration details.

The model predicts visual evolution rather than exposing a physics engine's state or task-success
scores. Its held-controller interface is suited to interactive exploration; custom-policy evaluation
requires a compatible action mapping and validation of the generated dynamics.

For runnable Python code, see the
[AgiBot simulation examples](https://github.com/reactor-team/reactor-cookbook/tree/main/robotics/simulation/cosmos-bimanual)
in the Reactor Cookbook.

Already have a physics simulator? Use
[Cosmos3 Nano Sim Transfer](/robotics/cosmos/nano-sim-transfer/overview) to render that simulator's
depth observations instead.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.