Simon Willison upgraded his Claude Token Counter tool to compare token usage across different Claude models. The analysis reveals that Claude Opus 4.7 uses 1.46x more tokens than 4.6 for text and up to 3x more for high-resolution images, effectively making it 40% more expensive despite identical pricing. The tool now supports comparisons between Opus 4.7, 4.6, Sonnet 4.6, and Haiku 4.5 models.
Background
Claude models from Anthropic process text using tokenization, where input text is converted into tokens that affect both model performance and API pricing. Different models may use different tokenizers, impacting cost efficiency.
- Source
- Simon Willison
- Published
- Apr 20, 2026 at 08:50 AM
- Score
- 6.0 / 10