Combined Claude + Codex token usage and cost tracking CLI
Project description
AI Usage Tracker
A comprehensive token usage and cost tracking tool for Claude Code and Codex CLI.
Overview
This tool combines usage data from both ccusage (Claude) and ccusage-codex (Codex) to provide a unified view of your AI token consumption and associated costs.
Quick Start
# Clone the repository
git clone https://github.com/jleechanorg/ai-usage-tracker.git
cd ai-usage-tracker
# Install dependencies (ccusage and ccusage-codex)
./scripts/install-dependencies.sh
# Run the report
python3 scripts/combined-usage-report.py
# Or install as a Claude skill
ln -s $(pwd)/.claude/skills/combined-usage ~/.claude/skills/combined-usage
# Then use: /combined-usage
Features
- Combined Reporting: View Claude and Codex usage side-by-side
- Cost Analysis: Track spending across both platforms
- Daily Averages: Calculate average usage over time
- Cache Efficiency: Monitor cache read rates and savings
- Flexible Output: Table format for humans, JSON for automation
- Claude Skill: Invoke directly from Claude Code CLI
Installation
Prerequisites
- Python 3.6 or higher
- Node.js and npm (for installing ccusage tools)
Automated Installation
We provide an automated installation script that installs both ccusage and ccusage-codex:
# Clone the repository
git clone https://github.com/jleechanorg/ai-usage-tracker.git
cd ai-usage-tracker
# Run the installation script
./scripts/install-dependencies.sh
The script will:
- Check if Node.js and npm are installed
- Install
ccusageglobally via npm - Install
ccusage-codexglobally via npm - Verify both installations
- Prompt before reinstalling if already installed
Manual Installation
If you prefer to install manually:
# Install ccusage
npm install -g ccusage
# Install ccusage-codex
npm install -g ccusage-codex
Additional Setup
- (Optional) Install as a Claude skill by symlinking:
ln -s $(pwd)/.claude/skills/combined-usage ~/.claude/skills/combined-usage
Usage
Command Line
# Default: Last 7 days
python3 scripts/combined-usage-report.py
# Custom date range
python3 scripts/combined-usage-report.py --days 14
# JSON output
python3 scripts/combined-usage-report.py --json
Claude Skill
If installed as a skill:
/combined-usage
Example Output
╔════════════╦═══════════════╦════════════╦═══════════════╦════════════╦═══════════════╦════════════╗
║ Date ║ Claude Tokens ║ Claude $ ║ Codex Tokens ║ Codex $ ║ Total Tokens ║ Total $ ║
╠════════════╬═══════════════╬════════════╬═══════════════╬════════════╬═══════════════╬════════════╣
║ 02-11 ║ 88,155,247 ║ $ 37.31 ║ 67,371,956 ║ $ 12.82 ║ 155,527,203 ║ $ 50.13 ║
║ 02-10 ║ 368,293,666 ║ $ 160.69 ║ 764,117,407 ║ $ 144.86 ║ 1,132,411,073 ║ $ 305.55 ║
...
╚════════════╩═══════════════╩════════════╩═══════════════╩════════════╩═══════════════╩════════════╝
DAILY AVERAGES (Last 7 complete days)
Claude: 237,410,924 tokens/day | $123.39/day
Codex: 511,863,756 tokens/day | $98.76/day
TOTAL: 749,274,679 tokens/day | $222.15/day
Documentation
See docs/usage-report.md for a complete example report with insights and recommendations.
Repository Structure
ai-usage-tracker/
├── README.md # This file
├── CHANGELOG.md # Version history
├── docs/
│ └── usage-report.md # Example report with analysis
├── scripts/
│ ├── install-dependencies.sh # Automated installation script
│ └── combined-usage-report.py # Main report generator
└── .claude/
└── skills/
└── combined-usage # Claude skill definition
Key Insights
Based on typical usage patterns:
- Token Distribution: Codex typically accounts for ~67% of total tokens
- Cost Ratio: Claude tokens are ~2.8x more expensive than Codex tokens
- Cache Efficiency: Both tools maintain 90%+ cache read rates
- Average Daily Cost: Typically ranges from $150-250/day for heavy usage
Contributing
Contributions welcome! Please feel free to submit issues or pull requests.
License
MIT License - Feel free to use and modify as needed.
Credits
Built for tracking usage across:
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file ai_usage_tracker-0.1.0.tar.gz.
File metadata
- Download URL: ai_usage_tracker-0.1.0.tar.gz
- Upload date:
- Size: 5.6 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
7c010db2d6a5a088ffdfeb07b6246126fc25252d55f29de642101d9a8c41ff8c
|
|
| MD5 |
a9140889d6dc58c0e5427f0d13f91000
|
|
| BLAKE2b-256 |
e78f3aaeb45f99ae01ce958ae21c2d55b6c076bd0c3c2adf7eff2809f00e5571
|
File details
Details for the file ai_usage_tracker-0.1.0-py3-none-any.whl.
File metadata
- Download URL: ai_usage_tracker-0.1.0-py3-none-any.whl
- Upload date:
- Size: 6.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
8aee9178664ab7a265a65085152908a03bc8a768def11f5f6cc85488640b70e1
|
|
| MD5 |
a5cd8562d66fc270b3289c74dd35f4fa
|
|
| BLAKE2b-256 |
dba4a6185e84c118b2f3fca323eaf33e507aeb29fa8ce501e6751f94e90481b1
|