Petal Labs Documentation

Petal Labs Documentation

Ship intelligent workflows without losing reliability.

Iris gives you a unified Go SDK for LLM providers. Petal Flow turns those models into durable, inspectable graphs. Cortex adds persistent memory and knowledge retrieval. PetalTrace provides deep observability into agent execution.

Iris SDK Fluent client, streaming, tools, embeddings, images, and a polished CLI.

Petal Flow Orchestrate graphs, routers, guards, and human-in-the-loop steps.

Cortex Memory service with conversation history, knowledge store, and entity tracking.

PetalTrace Agent observability with full prompt capture, cost tracking, and run replay.

Get Started

Choose your path

Iris Overview Learn how Iris unifies providers, offers streaming, and powers multimodal apps.

Petal Flow Overview See how graphs, nodes, and runtimes turn Iris calls into repeatable workflows.

Cortex Overview Add persistent memory, knowledge retrieval, and entity tracking to your agents.

PetalTrace Overview Capture LLM prompts, track costs, compare runs, and enable agent self-inspection.

Need the source?

Browse the repositories on GitHub:

• Iris SDK
• PetalFlow
• Cortex
• PetalTrace