LLM Pricing Calculator
Compare API prices and calculate the cost of your workload.
models.dev · Fetched
OpenRouter · Fetched
Feeds are checked hourly while this page is used. Saved snapshots older than seven days are withheld.
The total includes one cold request and the remaining requests at the selected cache-hit share.
Showing 600 of 8,806 API routes. A model served by different providers appears separately. Prices are USD per 1M text tokens; the displayed rate uses your input length. A $0 listed token rate can still require a paid plan or self-hosting.
Scroll the table horizontally to see rates and estimates.
What the estimate includes
Input, visible output, additional reasoning tokens, published cache reads/writes, and published OpenRouter request charges. Reasoning tokens are added at the listed reasoning rate, or the output rate when a separate rate is absent. Keep reasoning tokens separate from visible output to avoid counting them twice.
Caching assumes one cold request followed by hits on the chosen repeated input share. Published cache-write rates replace the regular input rate for newly cached tokens. Unlisted write charges, cache storage, taxes, credit purchase fees, subscriptions, batch discounts absent from the selected route, tools/search, and image/audio/video charges are excluded. Input modality filters show compatibility; these remain text-token estimates.
Unknown required rates prevent an estimate. Known context and output limits are checked; missing limits and provider access restrictions still need checking. Sources are external catalogs, and “fetched” records when we retrieved them rather than when a provider last verified a price.