How to use vertex ai with autogen?

[ks885522]

Hey guys,
I got this error

ModuleNotFoundError: No module named 'autogen.gemini' 

Is anyone know that how to install autogen.gemini dependency in python?

I have tried several way like

pip install autogen-gemini pip install autogen-agentchat[gemini] pip install pyautogen[gemini] 

I just want to try use vertex ai with autogen
here is my model config

gemini_2_5_flash_lite_us_native: provider: autogen.gemini.GoogleVertexAIClient config: # 這是為了滿足 (C) 您的 load_component 需求 model_client_cls: autogen.gemini.GoogleVertexAIClient # 這是 (B) 和 (A) 中 Vertex AI 的模型 ID model: "gemini-2.5-flash-lite" # 這是 (A) 中 Vertex AI 的必要參數 project_id: ${VERTEXAI_PROJECT} location: ${VERTEXAI_LOCATION} # 3. model_info (建議填寫)： model_info: family: gemini-2.5-flash-lite function_calling: true json_output: true vision: true structured_output: true 

[KostiushkoOleksandr]

The autogen.gemini module isn’t published as a standalone package — that’s why pip can’t find it.
It’s part of the Microsoft AutoGen project, but the Gemini client hasn’t been released yet.

Try installing the latest development version directly from GitHub:
pip install "git+https://github.com/microsoft/autogen.git"

[ks885522]

Many thanks I will try it and keep tracking when it will release