Releases: KurtWeston/token-count
Releases · KurtWeston/token-count
token-count v1.0.0
🚀 Token Count v1.0.0 - Initial Release
A powerful Rust CLI tool for calculating token counts across multiple LLM tokenizers. Perfect for estimating API costs, managing context limits, and optimizing prompts for GPT, Claude, and Llama models.
✨ Key Features
- Multi-Model Support: Compatible with GPT-3.5, GPT-4, Claude 2/3, and Llama 2/3 tokenizers
- Flexible Input: Read from files, stdin, or command-line arguments
- Cost Estimation: Real-time API cost calculations based on current pricing
- Batch Processing: Analyze multiple files simultaneously
- Interactive Mode: Test prompts with real-time token counting
- Model Comparison: Compare token counts across different models side-by-side
- Efficiency Metrics: View token-to-character ratios and encoding efficiency
- JSON Output: Programmatic integration support
- Colorized Output: Clear, readable cost breakdowns per model