# Sinapsis Product — different tools, one architecture

> Compose models, data, APIs and custom Python code into AI systems that can run on different infrastructure and evolve without being rebuilt.

## Features

- **Reusable templates:** Wrap models, libraries, APIs or Python code in templates reusable across projects.
- **Composable agents:** Connect templates into agents with YAML, Python or the visual editor.
- **One data contract:** Use DataContainer for images, audio, text and tables, with Universal Data Transport between processes and machines.
- **Isolated runtimes:** Run components with different framework and dependency requirements side by side.
- **Independent scaling:** Scale individual workload steps with retries and validated configuration.
- **Integration ecosystem:** Public Python packages and custom adapters connect models, frameworks, data stores and services; actual supported operations vary by package.

## Deployment options

Sinapsis describes support for Sinapsis Cloud, your cloud, on-premises and edge environments. The public product page includes the README-aligned CLI example `uv run sinx run agent_config.yml`; refer to the technical documentation for current command support and installation.

## More information

- [Product page](https://sinapsis.tech/product)
- [Technical documentation](https://docs.sinapsis.tech/docs/latest)
- [Source code](https://github.com/SinapsisHub/sinapsis-hub)
- [Pricing](https://sinapsis.tech/pricing.md)
