Skip to main content
A key pays for every request from your credits, at the prices below, on every plan. A model’s price is its maker’s own list price. For some models the maker no longer sells, it is the maker’s last published list price. Where neither applies, or no host can serve a tier at the maker’s price, the price is 1.25 times the list price of the lowest-priced main host Morse uses for it. Where a maker’s rate is below every host’s cost, that rate rises to the lowest one a host can serve. When a request costs Morse less than its price, half of the difference returns to your balance as a rebate a few minutes after the hour it finished in ends. One credit is $0.01. GET /v1/models lists the same prices in micros, a millionth of a dollar. Morse writes this page from the prices production bills, so it changes when they do.

How chat is priced

Chat prices are per million tokens. Input is the prompt, and cached input is the part of it the model’s vendor reads from its prompt cache. Output counts every token the model writes, its reasoning included. Context is the most tokens a request can hold, prompt and output together. A model has one id, and a row for each privacy tier it is sold at. The price of a tier is the same whichever of its providers serves the request. The Providers column lists them, each with the precision it publishes for a chat model’s weights, or unknown where it publishes none. Morse chooses between them, and provider.quantizations limits a request to the providers that serve a precision it names. On anon and private, the provider that serves a request reads its prompt. On tee and e2ee, the attestation names the vendor. GET /v1/models lists the same tiers and prices for each model. A “Prompt over” row is the price of a long prompt. Once a prompt passes that length, the whole request, its output included, pays those prices. Before a chat starts, Morse holds the most it could cost: the prompt at the cache-write price, plus the output limit. Set a lower max_tokens to hold less. When the reply ends, Morse charges what the tokens cost and returns the rest. See Billing.

Images

Image generation is not offered to API keys for now. The Morse app still makes images.

Web search, page reading, research, and transcription

Web search, page reading, research, and transcription are not offered to API keys for now. The Morse app still uses them.

Every price as a table