LLM Pricing Calculator (www.llm-prices.com)

A simple web-based LLM pricing calculator has appeared to help teams estimate inference costs by letting you plug in prompt (input) tokens, cached input tokens, and completion (output) tokens along with per‑million‑token rates. It separately accounts for cached vs uncached input tokens and output tokens, displays a total cost (in dollars and cents) and provides model-specific per‑million token prices you can select from. The tool is open‑source (GitHub link included) and notes that tokenizers differ between models, so token‑based comparisons aren’t perfectly exact; the interface also appears to pull live price data. For AI/ML practitioners and product teams this is a useful, practical utility for budgeting and architecture tradeoffs. It makes clear how prompt length, caching strategies (e.g., reusing encoded context in RAG), and output token volume drive costs, and lets you compare models on a per‑token basis to inform choices like aggressive prompt truncation, batching, or increasing cache hit rates. The explicit separation of cached inputs and outputs highlights cost levers often overlooked during prototyping, helping teams optimize latency/cost tradeoffs and forecast monthly spend more accurately.