Agent with Thinking Budget

Cap Gemini 2.5 Pro’s reasoning with thinking_budget and surface thought summaries with include_thoughts.

An example of how to use the thinking budget parameter with the Gemini model. This requires google-genai > 1.10.0

Google has scheduled gemini-2.5-pro for shutdown on October 16, 2026 and recommends gemini-3.1-pro-preview. Gemini 3 models use thinking_level instead of a token thinking_budget. See Gemini deprecations and Gemini thinking.

agent_with_thinking_budget.py

"""
An example of how to use the thinking budget parameter with the Gemini model.
This requires `google-genai > 1.10.0`

- Turn off thinking use thinking_budget=0
- Turn on dynamic thinking use thinking_budget=-1
- To use a specific thinking token budget (e.g. 1280) use thinking_budget=1280
- Use include_thoughts=True to get the thought summaries in the response.
"""

from agno.agent import Agent
from agno.models.google import Gemini

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

task = (
    "Three missionaries and three cannibals need to cross a river. "
    "They have a boat that can carry up to two people at a time. "
    "If, at any time, the cannibals outnumber the missionaries on either side of the river, the cannibals will eat the missionaries. "
    "How can all six people get across the river safely? Provide a step-by-step solution and show the solutions as an ascii diagram"
)

agent = Agent(
    model=Gemini(id="gemini-2.5-pro", thinking_budget=1280, include_thoughts=True),
    markdown=True,
)
agent.print_response(task, stream=True)

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

if __name__ == "__main__":
    pass

Run the Example

1. Set up your virtual environment

   Mac

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

   uv venv --python 3.12
   .venv\Scripts\activate
   
2. Install dependencies

   uv pip install -U agno google-genai
   
3. Export your Google API key

   Mac/Linux

   export GOOGLE_API_KEY="your_google_api_key_here"
   

   $Env:GOOGLE_API_KEY="your_google_api_key_here"
   
4. Migrate the Gemini thinking configuration

   Replace Gemini(id="gemini-2.5-pro", thinking_budget=1280, include_thoughts=True) with Gemini(id="gemini-3.1-pro-preview", thinking_level="low", include_thoughts=True) in the saved file.

5. Run the example

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

   python agent_with_thinking_budget.py
   

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