Google PDF Input Local

Attach a local PDF to a Gemini agent and ask follow-up questions with history.

pdf_input_local.py
"""
Google Pdf Input Local
======================

Cookbook example for `google/gemini/pdf_input_local.py`.
"""

from pathlib import Path

from agno.agent import Agent
from agno.media import File
from agno.models.google import Gemini
from agno.utils.media import download_file

# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------

pdf_path = Path(__file__).parent.joinpath("ThaiRecipes.pdf")

# Download the file using the download_file function
download_file(
    "https://agno-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf", str(pdf_path)
)

agent = Agent(
    model=Gemini(id="gemini-3.7-flash"),
    markdown=True,
    add_history_to_context=True,
)

agent.print_response(
    "Summarize the contents of the attached file.",
    files=[File(filepath=pdf_path)],
)
agent.print_response("Suggest me a recipe from the attached file.")

# ---------------------------------------------------------------------------
# Run Agent
# ---------------------------------------------------------------------------

if __name__ == "__main__":
    pass

Run the Example

Set up your virtual environment

uv venv --python 3.12
source .venv/bin/activate

Install dependencies

uv pip install -U agno google-genai

Export your Google API key

export GOOGLE_API_KEY="your_google_api_key_here"

Keep the PDF in conversation history

Add from agno.db.in_memory import InMemoryDb and db=InMemoryDb() to the agent configuration. This keeps the first PDF and response available to the follow-up within this process. Use a durable database for history across restarts.

Run the example

Save the code above as pdf_input_local.py, then run:

python pdf_input_local.py

Full source: cookbook/90_models/google/gemini/pdf_input_local.py