Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.egisai.co/llms.txt

Use this file to discover all available pages before exploring further.

egisai is a regular Python package. Install it with pip (or your packaging tool of choice — Poetry, uv, Hatch all work the same way) into the same environment that runs your AI application.

​
Requirements

ComponentRequirement
Python3.11 or newer
Operating systemLinux, macOS, Windows (any platform supported by CPython)
NetworkOutbound HTTPS to the EgisAI control plane and to your LLM provider(s)
The SDK has minimal direct dependencies. Provider integrations are activated only if the matching SDK is importable at runtime, so you do not pay for what you don’t use.

​
Install with pip

pip install "egisai[all]"
The core only install is appropriate when you intend to govern HTTP-level calls via httpx / requests and don’t use the official provider SDKs at all. See HTTP clients.

​
Install in a virtual environment

python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install --upgrade pip
pip install "egisai[all]"

​
Pinned versions

For reproducible builds, pin the SDK in your requirements.txt / pyproject.toml:
requirements.txt
egisai==0.10.0
openai>=1.40
anthropic>=0.40
google-generativeai>=0.8
Refer to the PyPI page for the latest released version.

​
Verify the install

python -c "import egisai; print(egisai.__version__)"
You should see the installed version, for example 0.10.0.

​
Upgrade

pip install --upgrade "egisai[all]"
The changelog describes user-facing changes per release.

​
Uninstall

pip uninstall egisai
Uninstalling removes the SDK; your provider SDKs and application code are unaffected. Any data already audited to your dashboard remains.

​
What’s next

Quickstart

Make your first governed call.

Configuration

Read every option egisai.init() accepts.