​
Different LLMs

AG2 installs OpenAI package by default. To use LLMs by other providers, you can install the following packages:

pip install autogen[gemini,anthropic,mistral,together,groq,cohere]

Check out the notebook and blogpost for more details.

​
LLM Caching

To use LLM caching with Redis, you need to install the Python package with the option redis:

pip install "autogen[redis]"

See LLM Caching for details.

​
IPython Code Executor

To use the IPython code executor, you need to install the jupyter-client and ipykernel packages:

pip install "autogen[ipython]"

To use the IPython code executor:

from autogen import UserProxyAgent

proxy = UserProxyAgent(name="proxy", code_execution_config={"executor": "ipython-embedded"})

​
blendsearch

pyautogen<0.2 offers a cost-effective hyperparameter optimization technique EcoOptiGen for tuning Large Language Models. Please install with the [blendsearch] option to use it.

pip install "autogen[blendsearch]<0.2"

Checkout Optimize for Code Generation and Optimize for Math for details.

​
retrievechat

autogen supports retrieval-augmented generation tasks such as question answering and code generation with RAG agents. Please install with the [retrievechat] option to use it with ChromaDB.

pip install "autogen[retrievechat]"

Alternatively autogen also supports PGVector and Qdrant which can be installed in place of ChromaDB, or alongside it.

pip install "autogen[retrievechat-pgvector]"

pip install "autogen[retrievechat-qdrant]"

RetrieveChat can handle various types of documents. By default, it can process plain text and PDF files, including formats such as ‘txt’, ‘json’, ‘csv’, ‘tsv’, ‘md’, ‘html’, ‘htm’, ‘rtf’, ‘rst’, ‘jsonl’, ‘log’, ‘xml’, ‘yaml’, ‘yml’ and ‘pdf’. If you install unstructured (pip install "unstructured[all-docs]"), additional document types such as ‘docx’, ‘doc’, ‘odt’, ‘pptx’, ‘ppt’, ‘xlsx’, ‘eml’, ‘msg’, ‘epub’ will also be supported.

You can find a list of all supported document types by using autogen.retrieve_utils.TEXT_FORMATS.

Example notebooks:

Automated Code Generation and Question Answering with Retrieval Augmented Agents

Group Chat with Retrieval Augmented Generation (with 5 group member agents and 1 manager agent)

Automated Code Generation and Question Answering with Qdrant based Retrieval Augmented Agents

​
Teachability

To use Teachability, please install AG2 with the [teachable] option.

pip install "autogen[teachable]"

Example notebook: Chatting with a teachable agent

​
Large Multimodal Model (LMM) Agents

We offered Multimodal Conversable Agent and LLaVA Agent. Please install with the [lmm] option to use it.

pip install "autogen[lmm]"

Example notebook: LLaVA Agent

​
mathchat

pyautogen<0.2 offers an experimental agent for math problem solving. Please install with the [mathchat] option to use it.

pip install "autogen[mathchat]<0.2"

Example notebook: Using MathChat to Solve Math Problems

​
Graph

To use a graph in GroupChat, particularly for graph visualization, please install AutoGen with the [graph] option.

pip install "autogen[graph]"

Example notebook: Finite State Machine graphs to set speaker transition constraints

​
Long Context Handling

AG2 includes support for handling long textual contexts by leveraging the LLMLingua library for text compression. To enable this functionality, please install AutoGen with the [long-context] option:

pip install "autogen[long-context]"
DockerIntroduction to AutoGen