Repolode
Submit
dynamiq-ai

dynamiq-ai/dynamiq

Python orchestration framework for building agentic AI and LLM applications, with support for RAG pipelines, agents, and multi-step workflows.

View on GitHubWebsite
Stars
1,073
Forks
154
Open issues
17
Last commit
yesterday

What is it?

Dynamiq is a Python framework for building applications on large language models. It provides building blocks for LLM calls, tool-using agents, retrieval-augmented generation (RAG) pipelines, and workflows that connect these nodes into a directed graph.

The framework covers document ingestion, including PDF, Word, PowerPoint and Excel reading, plus chunking, embedding and writing to vector stores such as Pinecone, Qdrant, Milvus, Weaviate and pgvector. Agents can call built-in tools, including code interpreters in E2B and Daytona sandboxes, and other agents can be used as tools for multi-agent setups. A graph orchestrator lets developers define custom state machines with conditional routing, and a memory module stores conversation history.

Install with pip install dynamiq. Optional extras such as vectorstores, files and sql keep heavier integration SDKs out of the base install.

Key features

  • 01Agents that call tools, including other agents, with a ReAct-style loop.
  • 02Workflows that run nodes in parallel or sequentially based on dependencies.
  • 03RAG building blocks for document conversion, splitting, embedding and retrieval.
  • 04Connectors for many LLM providers and vector databases.
  • 05Graph orchestrator for custom state machines with conditional edges.
  • 06Conversation memory backends for chatbots.

Use cases

  • Building a question-answering system over PDF or Office documents using a vector store.
  • Creating a multi-agent pipeline where a manager agent delegates research and writing tasks.
  • Adding code execution in sandboxes to an LLM agent.
  • Prototyping chatbots that keep conversation history.

Quick start

  1. 1

    Install Python 3.11 or later, then install the package with pip

    pip install dynamiq
  2. 2

    Or install with extras such as vector stores and file support

    pip install "dynamiq[vectorstores,files]"

Things to know

  • Requires Python 3.11 or later.
  • Some integrations need extra packages installed separately, such as the vectorstores or sql extras.
  • Running examples requires API keys for LLM providers such as OpenAI.

In collections

Open-source AI agent frameworks & apps

Topics

#llm#agents#rag#workflow-orchestration#vector-database#python#ai#generative-ai#gpt#llmops

FAQ

What is dynamiq?
Python orchestration framework for building agentic AI and LLM applications, with support for RAG pipelines, agents, and multi-step workflows. Dynamiq is a Python framework for building applications on large language models. It provides building blocks for LLM calls, tool-using agents, retrieval-augmented generation (RAG) pipelines, and workflows that connect these nodes into a directed graph.
Is dynamiq free to use? What license does it have?
dynamiq is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is dynamiq still maintained?
Actively maintained. The last commit to dynamiq was on Nov 6, 2026.
What can I use dynamiq for?
Building a question-answering system over PDF or Office documents using a vector store. Creating a multi-agent pipeline where a manager agent delegates research and writing tasks. Adding code execution in sandboxes to an LLM agent. Prototyping chatbots that keep conversation history.
How do I install dynamiq?
Install Python 3.11 or later, then install the package with pip: pip install dynamiq Or install with extras such as vector stores and file support: pip install "dynamiq[vectorstores,files]"
Where can I get dynamiq?
The source code and installation instructions are on GitHub: https://github.com/dynamiq-ai/dynamiq

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects