LLM Cost Calculator

A pure client-side calculator that estimates what a set of LLM API requests would cost. You add one card per request, pick a model from the built-in catalog (15 models across OpenAI, Anthropic, Google, DeepSeek, and Meta), enter input and output token counts, and the tool shows per-request cost and a running total.

1. On each request card, choose a model from the Select dropdown (labeled [Provider] Model). 2. Enter Input Tokens and Output Tokens (non-negative integers). 3. Click Add Request to add another card, or the trash icon on a card to remove it (only shown when more than one card exists). 4. Read each card's cost in the header and the Total Cost in the footer bar.

Per-request cost = (inputTokens / 1_000_000 * model.inputPrice) + (outputTokens / 1_000_000 * model.outputPrice), where inputPrice and outputPrice are USD per million tokens from the catalog. Total cost is the sum across all request cards. The formatCost helper pads small values: under $0.01 it shows 6 decimal places, under $1 four decimals, otherwise two decimals. A read-only Price Table at the bottom lists every model's per-million prices and context window.