HyperFlow workflows are built from nodes connected by data flows and process flows. Nodes are the functional building blocks that perform specific operations like content processing, AI generation, data transformation, and flow control. Each node exposes input ports for receiving data, output ports for producing results, and exit ports for directing process flow to subsequent nodes.
Super nodes provide access to multiple services that implement the node’s core functionality through different approaches or providers. For example, the Transform Content super node offers access to PDF transformation services like LlamaParse, Mistral OCR, and HyperFlow LLM PDF Transformer - each providing document transformation capabilities with different specializations and trade-offs. Services make HyperFlow's capabilities extensible, allowing new implementations to be added without changing the core node interface or containing flow-graph.
Both nodes and services are highly configurable through parameters that control their behavior, from simple settings like temperature and token limits to complex configurations like embedding models and transformation modes. The strongly-typed Data Value system ensures reliable data flow between nodes while supporting automatic type conversion when compatible types are connected. This combination of flexible nodes, extensible services, and robust data flow creates a powerful platform for building sophisticated AI applications.
Complete documentation for all HyperFlow nodes, including parameters, input/output ports, common workflows, and usage examples.
Detailed documentation for all available services, covering configuration parameters, capabilities, best practices, and integration guidance.
Comprehensive guide to HyperFlow's strongly-typed data system, including type compatibility rules and automatic conversions.