AI
Tokens, model pricing and what a workload actually costs to run.
4 calculators
- LLM Token Cost CalculatorModel pricing is quoted per million tokens and input almost always costs less than output. Enter the two prices from whichever provider you use.
- Tokens to Words CalculatorTokens are pieces of words, not words. This converts in both directions and tells you whether what you have fits in a context window.
- GPU VRAM CalculatorThe weights are only part of it. The context cache grows with how much you feed the model, and it is what usually pushes a model off a card.
- AI API Monthly Cost CalculatorBuilt for anyone putting an AI feature in a product. The number that decides whether it works is not the monthly total, it is the cost per user against what you charge.
Estimating AI costs
AI models charge by the token, a piece of a word. In English, one token is about three quarters of a word, so 1,000 tokens is roughly 750 words. The tokens to words calculator converts between the two, and the token cost calculator turns a token count into a price.
Input and output tokens are priced separately, and output usually costs several times more. The monthly API cost calculator multiplies a single request by how many you expect, which is where small costs per request become large bills.
The GPU VRAM calculator estimates the memory needed to run a model on your own hardware, from its size and precision. Prices change often, so compare the result with the current price list of the provider you use.