Develop Apps with Neat

What Neat Is

SiMa.ai Neat is an application-development framework for building and running AI applications on the SiMa.ai platform. It provides developers a set of Python and C++ APIs to execute and test compiled model artifacts (tar.gz models), compose AI applications that leverage the SoC's hardware blocks, and manage runtime execution.

In the broader SiMa.ai software ecosystem, Neat sits at the application layer, building on the SiMa.ai runtime stack and using GStreamer-based execution underneath so developers can stay focused on application logic instead of manually stitching together lower-level runtime pieces.

C++ or PyNeat?

Neat ships the same workflow through two front ends, so you can pick the one that fits your application:

• PyNeat — the Python bindings (pyneat). Best for quick iteration, notebooks, data-science workflows, and validating directly on the DevKit, where the Python runtime executes.
• C++ — the native simaai::neat API. Best for larger applications, tight integration with existing C++ codebases, and cross-compiled host-to-DevKit workflows.

Both target the same compiled model artifacts and the same DevKit runtime; the concepts and pages below apply to either.

Requirements

Before building applications, complete the Getting Started setup:

• Install and sync — install the Neat Library in your Palette SDK or Modalix environment, then pair and sync your DevKit if you are developing from a host workflow.
• Model artifact — use a precompiled model from the Model Zoo or compile your own model into a Modalix-ready archive.
• Runtime target — decide whether you want to iterate directly on Modalix with PyNeat or cross-compile a C++ application from Palette SDK.

See Installation, Pair a DevKit, and Compile a Model if any of those steps are not ready yet.

How It Works

Neat gives you a direct mental model for that path. A compiled model archive (.tar.gz) becomes a Model component, application logic is assembled as a Graph, and that graph is built and executed as a Run object on the SoC. The same workflow is designed to work well with agentic development too, so teams can explore, build, and iterate faster.

The Hello Neat! pages help you run your first inference, the Development Workflow pages explain the main concepts in more detail, and the tutorials show how to apply them to real application patterns.

Start Here

Start from a working environment and build up the core Neat application workflow.

• Hello Neat!Run a minimal Neat application and verify the development loop.
• Development WorkflowLearn the `Model`, `Graph`, and `Run` workflow in more detail.
• TutorialsFollow guided examples that walk through real Neat application patterns.

Build More

Use these sections when you are ready to build richer applications or inspect the API surface.

• Advanced ConceptsUnderstand graphs, formats, memory, threading, and runtime behavior.
• ReferenceBrowse C++, Python, Model Compiler, troubleshooting, and supporting material.
• ContributeUnderstand architecture, source builds, testing expectations, and repo conventions.