Skip to main content

Installing Dhenara AI

Dhenara AI is available on PyPI and can be installed using pip or uv.

Requirements​

  • Python 3.13

    Dhenara currently targets Python 3.13 to keep typing, providers, and structured-output support consistent.

Basic Installation​

pip install dhenara-ai

This installs the public dhenara-ai package with its provider integrations and typed client surfaces.

uv supports two common workflows:

A) You have a project (pyproject.toml)​

# (optional) create a new project
uv init

# add dependency to pyproject.toml
uv add dhenara-ai

# create/update .venv and install from the lockfile
uv sync

Run without activating the venv:

uv run python -c "import dhenara.ai as dai; print(dai.__version__)"

B) You just want a quick install into an existing venv​

uv pip install dhenara-ai

It's best practice to use a virtual environment for Python projects:

# Create a virtual environment
python -m venv .venv

# Activate the environment (Linux/Mac)
source .venv/bin/activate

# Activate the environment (Windows)
.venv\Scripts\activate

# Install Dhenara
pip install dhenara-ai

Using uv:

# Create a virtual environment (creates .venv)
uv venv

# Option 1: activate + use normal python/pip
source .venv/bin/activate
pip install dhenara-ai

# Option 2: don't activate; run/install via uv
uv pip install dhenara-ai
uv run python -c "import dhenara.ai as dai; print(dai.__version__)"

Installing from Source​

git clone https://github.com/dhenara/dhenara-ai.git
cd dhenara-ai

# Editable install
pip install -e .

# Or with uv
uv sync --all-extras

Provider-specific Dependencies​

Dhenara automatically installs the necessary dependencies for all supported providers.

Provider-specific Requirements​

Different AI providers have specific requirements:

  • OpenAI: Requires an OpenAI API key
  • Google Gemini Developer API: Requires a Gemini API key under the google_gemini_api provider block
  • Google Vertex AI: Requires service account credentials
  • Anthropic: Requires an Anthropic API key
  • Microsoft OpenAI: Requires an API key plus an Azure OpenAI or Microsoft Foundry OpenAI v1 endpoint
  • Amazon Bedrock: Requires AWS credentials with Bedrock access

For Microsoft-hosted OpenAI-compatible deployments, use the microsoft_openai provider block. The older microsoft_azure_ai text-generation surface is not the recommended public setup.

Credential Discovery​

The recommended public setup uses ResourceConfig and one of these two flows:

  1. Pass an explicit credentials file path to ResourceConfig.load_from_file().
  2. Set DAI_SECRET_CONFIG_DIR and place dai_credentials.yaml inside that directory.

If DAI_SECRET_CONFIG_DIR is unset, Dhenara AI falls back to /run/secrets/dai/dai_credentials.yaml.

See Resource Configuration for the current credentials-file shape.

Verifying Installation​

You can verify your installation with:

import dhenara.ai as dai

print(f"Dhenara version: {dai.__version__}")

Troubleshooting​

If you encounter issues with installation:

  • Ensure you have the latest pip: pip install --upgrade pip
  • Try installing with the --verbose flag: pip install --verbose dhenara-ai
  • For provider-specific issues, check that you have the required credentials properly configured

Next Steps​

  • Continue to Quick Start for your first Dhenara application