The AI Cost Calculator that checked its own numbers: GPT-5.6, Claude, and Gemini pricing pages all changed in the last month.
Anthropic quietly made an “introductory” rate permanent. Assumes a flat cache-hit rate applied to input tokens only, at each model’s published cached-input rate. Output/”thinking” tokens are never cached.
Key takeaways
- Try Office Apps: What the sticker price doesn’t show you. Every calculator, including this one, reduces a real bill to two numbers: input rate times…
- Those tokens are billed at the model’s standard output rate and count against context limits, but they’re not returned in the response.
- Learn AI Tools: Why do output tokens cost more than input tokens?
Do thinking or reasoning tokens cost extra?
Read the full article: https://www.ainvasion.com/ai-cost-calculator/
Top comments (0)