# Sinapsis — A single contract for production AI systems > Sinapsis is an open-source, Python-native AI framework and a separately hosted platform for building end-to-end AI agents, workflows and multimodal applications. A shared computational contract connects otherwise fragmented models, libraries, APIs, data pipelines and custom code, enabling reusable components, provider choice and consistent execution across heterogeneous environments. Sinapsis is designed for teams building **production-ready AI applications**, not only chatbots. Its goal is to reduce integration glue between LLMs, vision, speech, document processing, databases, training and inference. The architecture separates *what the application does* (the Agent graph and reusable Templates) from *where and how it runs* (the runtime and infrastructure). Use a managed Sinapsis Cloud subscription for hosted services; the Free option is self-managed use of eligible open-source packages without managed hosting. The current Sinapsis Hub repository states that its open-source components are available under your choice of MIT or Apache License 2.0 unless otherwise noted. Individual packages and historical releases may use other licences. ## Core architecture and single contract - [How the Sinapsis contract works](https://sinapsis.tech/architecture/index.md): Agent graph, reusable Templates, DataContainer and typed DataPackets, Universal Data Transport, runtime execution and the boundaries between them. Static, crawler-accessible summary of key documentation concepts. - [Why choose Sinapsis?](https://sinapsis.tech/why-sinapsis/index.md): Reuse, interoperability, multimodal workflows, freedom of provider and runtime choice, reduced integration overhead and practical trade-offs. - [End-to-end production AI](https://sinapsis.tech/production-ai/index.md): Deployment, isolation, independent scaling, observability, autoscaling and one-click hosting considerations. Distinguishes platform goals from guaranteed features and service levels. - [Supported AI ecosystem and public Python packages](https://sinapsis.tech/ecosystem/index.md): Public packages for OpenAI, Anthropic, vLLM, LangChain, Hugging Face, OCR, vision, speech, time series and data workflows, with links to package pages. Package versions and licensing vary. ## Framework comparisons and selection guidance - [Sinapsis vs LangChain and LangGraph](https://sinapsis.tech/compare-langchain/index.md): Language-model tooling and stateful graph orchestration versus a common multimodal AI computational contract; includes where they work together. - [Sinapsis vs n8n](https://sinapsis.tech/compare-n8n/index.md): Python-native AI computation and reusable components versus low-code business process automation; includes complementary use cases. ## Product, usage and pricing - [Sinapsis overview](https://sinapsis.tech/index.md): What Sinapsis is, intended users and paths to build with it. - [Product and deployment](https://sinapsis.tech/product.md): Templates, Agents, multimodal processing, runtime boundaries and platform positioning. - [Current plans](https://sinapsis.tech/pricing.md): Free open-source self-use without hosting; hosted Pro at £69.99 per named seat/month with 360,000 daily credits; hosted Team at £299.99/month for up to ten members with 360,000 credits per active member pooled each day; Enterprise by enquiry. Headline prices exclude VAT; checkout and Terms govern. ## Primary developer documentation and open source - [Developer documentation](https://docs.sinapsis.tech/docs/latest): Official implementation guides, concepts, APIs and tutorials. Some versioned documentation content is dynamically rendered; the static Sinapsis architecture and ecosystem resources above provide a plain-text entry point, not a replacement for API specifications. - [Concepts](https://docs.sinapsis.tech/docs/latest/concepts): Agents, Templates, DataContainers and Universal Data Transport. - [Architecture reference](https://docs.sinapsis.tech/docs/latest/architecture): Declarative execution, models and transport details. - [Tutorials](https://docs.sinapsis.tech/docs/latest/tutorials): Getting started and applied use cases. - [Current open-source hub](https://github.com/SinapsisHub/sinapsis-hub): Repository README, quick start and MIT/Apache-2.0 dual-licence statement (unless otherwise noted). - [Python package listing](https://pypi.org/project/sinapsis/): Historical PyPI release and metadata; its description may lag the current repository architecture and licensing. Read licence terms for the exact installed distribution. ## Official company and legal information - [Sinapsis Group Limited legal index](https://sinapsis.tech/legal/index.md): Legal documents for the hosted service. - [Terms of Service](https://sinapsis.tech/legal/terms/index.md): Hosted subscriptions, plan entitlements, limits, customer responsibilities and consumer rights. - [Privacy Policy](https://sinapsis.tech/legal/privacy/index.md): How Sinapsis Group Limited handles personal information. - [Trademark Policy](https://sinapsis.tech/legal/trademarks/index.md): Brand usage rights and restrictions. ## Optional - [Extended plain-text reference](https://sinapsis.tech/llms-full.txt): Concatenated architecture, ecosystem, comparison, production and pricing explanations, suitable for crawlers that cannot render dynamic docs. - [XML sitemap](https://sinapsis.tech/sitemap.xml): Crawlable HTML URLs. - [Contact](mailto:hello@sinapsis.tech): Sales and general enquiries.