llm-usage-tracker

A drop-in token and cost tracker for popular LLM libraries with caching awareness. Automatically tracks usage across multiple LLM providers by monkey-patching their API calls.

Features

• Zero-config tracking: Just import and start tracking
• Multi-provider support: OpenAI, LiteLLM, Google Gemini
• Cost calculation: Automatic cost computation based on current pricing
• Session management: Track usage across different sessions
• Caching awareness: Handles cached responses appropriately

Supported Libraries

• OpenAI v1 (client + module-level)
• OpenAI v0 (legacy ChatCompletion.create)
• LiteLLM (optional)
• Google google-generativeai (optional)

Installation

pip install llm-usage-tracker

# With optional dependencies:
pip install llm-usage-tracker[litellm]  # For LiteLLM support
pip install llm-usage-tracker[gemini]   # For Google Gemini support
pip install llm-usage-tracker[all]      # All optional dependencies

Quick Start

from llm_usage_tracker import setup_patch, print_usage_costs

# Enable tracking
setup_patch()

# Your existing OpenAI/LiteLLM/Gemini code works unchanged
import openai
client = openai.OpenAI()
response = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[{"role": "user", "content": "Hello!"}]
)

# Print usage summary
print_usage_costs(OPENAI_DEFAULT)

Usage with Sessions

from llm_usage_tracker import UsageSession, OPENAI_DEFAULT

with UsageSession(OPENAI_DEFAULT, label="my-session") as session:
    # Your LLM calls here
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": "Analyze this data"}]
    )

# Session results automatically saved
print(f"Session cost: ${session.result['grand_total_usd']:.4f}")

Advanced Features

Per-Agent Cost Tracking

from llm_usage_tracker import setup_patch, set_scope, print_usage_costs_by_scope, OPENAI_DEFAULT

setup_patch()

# Track costs per agent/component
set_scope("agent:researcher")
# ... LLM calls for research agent

set_scope("agent:writer")
# ... LLM calls for writer agent

# See breakdown by scope
print_usage_costs_by_scope(OPENAI_DEFAULT)

CrewAI Integration

from llm_usage_tracker.decorators import setup_agent_logging, setup_tool_logging
from llm_usage_tracker.trace_log import TraceLogger

# Automatic logging for CrewAI agents and tools
setup_agent_logging("Research Agent", tracer)(MyAgent)
setup_tool_logging("Web Search", tracer)(MyTool)

Use Cases

• Multi-Agent Systems: Track costs per agent in CrewAI/AutoGen workflows
• Production Monitoring: Monitor LLM costs in production applications
• Development: Understand token usage during development
• Cost Optimization: Identify expensive operations
• Billing/Reporting: Generate detailed usage reports

License

MIT