LLM API Cost Calculator

Estimate what a workload costs per month on current Claude, GPT, and Gemini models. Prices are list prices per million tokens as of August 2026 — always confirm against the vendor's pricing page before committing. Everything runs in your browser.

ModelRequests per monthAPI calls across the month.Average input tokens per requestSystem prompt + history + user message. ~4 characters ≈ 1 token.Average output tokens per requestThe response the model generates.Cached input share: 0%Portion of input served from prompt cache at ~0.1× the input price (a stable system prompt often makes this high).Batch API (−50%, for workloads that can wait up to ~24h)

Estimated monthly cost$1,350

Input: $600Output: $750Per request: $0.0135