Claude Token Counter Tool Updated with Model Comparison Feature
Original: Claude Token Counter, now with model comparisons
Why This Matters
Token inflation in new models affects AI deployment costs and pricing strategies
Simon Willison upgraded his Claude Token Counter tool to compare tokenization across different Claude models. Claude Opus 4.7's new tokenizer produces 1.0-1.35x more tokens than previous versions, making it approximately 40% more expensive despite same pricing per token.
The updated tool allows users to compare token counts across Claude models including Opus 4.7, 4.6, Sonnet 4.6, and Haiku 4.5. Claude Opus 4.7 introduced an updated tokenizer that processes text differently, resulting in 1.0-1.35x more tokens depending on content type. Testing the Opus 4.7 system prompt showed a 1.46x increase in tokens compared to 4.6. While pricing remains the same at $5 per million input tokens and $25 per million output tokens, the token inflation makes Opus 4.7 effectively 40% more expensive. For images, the tool showed a 3.01x increase in tokens for high-resolution images due to 4.7's ability to handle up to 2,576 pixels on the long edge, though similar-sized images show minimal differences. A PDF test showed only a 1.08x multiplier, significantly lower than raw text increases.