To add dependencies using the UV Package Manager, you can use it similarly to how you'd use pip, but faster and more efficiently. Here's how:

✅ Step-by-Step: Adding Dependencies with UV

1. Install a package (like pip)

uv pip install <package-name>

Example:

uv pip install numpy pandas

This installs numpy and pandas into your environment.

2. Create a requirements.txt or pyproject.toml (optional but recommended)

If you're managing a project, it's good practice to declare dependencies in one of these files.

3. Compile a lock file for reproducibility

uv pip compile requirements.in

This generates a requirements.lock file that freezes exact versions.

4. Install from a lock file (like pip-sync)

uv pip sync requirements.lock

Ensures your environment exactly matches the locked versions.

5. Using with pyproject.toml (for modern Python projects)

If you're using a pyproject.toml, you can let UV handle dependencies via standard PEP 621 configuration.

🔄 Example Workflow

# Step 1: Create a virtual environment
uv venv

# Step 2: Activate the environment
source .venv/bin/activate   # (Linux/macOS)
.venv\Scripts\activate      # (Windows)

# Step 3: Add dependencies
uv pip install fastapi uvicorn

# Step 4: (Optional) Freeze versions into a lock file
uv pip compile requirements.in

# Step 5: Reinstall in new environments
uv pip sync requirements.lock