CostPerPrompt, launched on Hacker News, provides live AI API pricing for major providers and a calculator that estimates costs based on real-world workload parameters like input tokens, output tokens, and caching. The tool aggregates current prices from providers such as OpenAI, Anthropic, and Google, updating as they change. It also includes a comparison feature that lets users evaluate multiple models side by side for specific use cases. The creator posted the site as a Show HN, inviting community feedback on accuracy and usability. As of August 2, 2026, the site is operational and free to use.
Pricing for AI APIs is a mess. Every provider slaps a different number on a different token. Some charge extra for caching, others for reasoning. Developers are left guessing. CostPerPrompt cuts through that noise. It turns a spreadsheet nightmare into a simple query. That is a quiet revolution.
We are moving from the era of AI hype to the era of AI budgets. Companies are asking: what does this actually cost? Tools like this make the answer visible. They force transparency. They let you compare, plan, and maybe even negotiate. The future is not about who has the smartest model. It is about who can deploy it without blowing the budget. CostPerPrompt is a compass for that journey.