> ## Documentation Index
> Fetch the complete documentation index at: https://phidatainc.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Advanced Compression

> Set a context token based limit for tool call compression.

```python advanced_compression.py theme={null}
"""
Advanced Compression
=============================

This example shows how to set a context token based limit for tool call compression.
"""

from agno.agent import Agent
from agno.compression.manager import CompressionManager
from agno.db.sqlite import SqliteDb
from agno.models.openai import OpenAIResponses
from agno.tools.websearch import WebSearchTools

compression_prompt = """
    You are a compression expert. Your goal is to compress web search results for a competitive intelligence analyst.
    
    YOUR GOAL: Extract only actionable competitive insights while being extremely concise.
    
    MUST PRESERVE:
    - Competitor names and specific actions (product launches, partnerships, acquisitions, pricing changes)
    - Exact numbers (revenue, market share, growth rates, pricing, headcount)
    - Precise dates (announcement dates, launch dates, deal dates)
    - Direct quotes from executives or official statements
    - Funding rounds and valuations
    
    MUST REMOVE:
    - Company history and background information
    - General industry trends (unless competitor-specific)
    - Analyst opinions and speculation (keep only facts)
    - Detailed product descriptions (keep only key differentiators and pricing)
    - Marketing fluff and promotional language
    
    OUTPUT FORMAT:
    Return a bullet-point list where each line follows this format:
    "[Company Name] - [Date]: [Action/Event] ([Key Numbers/Details])"
    
    Keep it under 200 words total. Be ruthlessly concise. Facts only.
    
    Example:
    - Acme Corp - Mar 15, 2024: Launched AcmeGPT at $99/user/month, targeting enterprise market
    - TechCo - Feb 10, 2024: Acquired DataStart for $150M, gaining 500 enterprise customers
"""

compression_manager = CompressionManager(
    model=OpenAIResponses(id="gpt-5-mini"),
    compress_token_limit=5000,
    compress_tool_call_instructions=compression_prompt,
)

# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------
agent = Agent(
    model=OpenAIResponses(id="gpt-5-mini"),
    tools=[WebSearchTools()],
    description="Specialized in tracking competitor activities",
    instructions="Use the search tools and always use the latest information and data.",
    db=SqliteDb(db_file="tmp/token_based_tool_call_compression.db"),
    compression_manager=compression_manager,
    add_history_to_context=True,  # Add history to context
    num_history_runs=3,
    session_id="token_based_tool_call_compression",
)

# ---------------------------------------------------------------------------
# Run Agent
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    agent.print_response(
        """
        Use the search tools and always use the latest information and data.
        Research recent activities (last 3 months) for these AI companies:
        
        1. OpenAI - product launches, partnerships, pricing
        2. Anthropic - new features, enterprise deals, funding
        3. Google DeepMind - research breakthroughs, product releases
        4. Meta AI - open source releases, research papers
       
        For each, find specific actions with dates and numbers.""",
        stream=True,
    )
```

## Run the Example

<Steps>
  <Snippet file="create-venv-step.mdx" />

  <Step title="Install dependencies">
    ```bash theme={null}
    uv pip install -U agno ddgs openai sqlalchemy
    ```
  </Step>

  <Step title="Export your OpenAI API key">
    <CodeGroup>
      ```bash Mac/Linux theme={null}
      export OPENAI_API_KEY="your_openai_api_key_here"
      ```

      ```bash Windows theme={null}
      $Env:OPENAI_API_KEY="your_openai_api_key_here"
      ```
    </CodeGroup>
  </Step>

  <Step title="Run the example">
    Save the code above as `advanced_compression.py`, then run:

    ```bash theme={null}
    python advanced_compression.py
    ```
  </Step>
</Steps>

Full source: [cookbook/02\_agents/14\_advanced/advanced\_compression.py](https://github.com/agno-agi/agno/blob/main/cookbook/02_agents/14_advanced/advanced_compression.py)