{"slug":"model-prices","domain":"research","title":"Model prices: every provider, cheapest first","summary":"The current version of every language model family, and what each provider charges for it per million tokens.","help":"model-prices: current language models and every provider's price, as JSON.\n\nRETURNS\n  One JSON object on stdout. `model-prices` with no arguments prints what it\n  takes.\n\nWHAT \"CURRENT\" MEANS\n  The newest version of each family. Once GLM 5.3 is out, GLM 5.2 is never\n  listed; each model names the versions it `supersedes`. A flash, mini or pro\n  tier is its own family, because it is its own product at its own price. A\n  family with no release in a year is left out unless `--max_age_days 0`.\n\nTHE LISTED PRICE IS ONE PROVIDER'S\n  An open-weight model can be served by forty providers at prices twenty times\n  apart. The price `frontier` lists is not the lowest; `providers` reads every\n  one. Before calling a price the best, check `quantization`: an fp4 endpoint\n  costs less than an fp8 one and can answer worse.\n\nSPEEDS ARE NOT PROVIDERS\n  A closed model costs the same at every host; its maker sells speeds. Each\n  offer says its `speed` — `standard`, `flex` (half price, slower), `fast`,\n  `priority`, `ultrafast` — and only standard offers are counted as the\n  cheapest. `cheapest_flex` is the price to wait. An offer's `long_prompt` is\n  what it charges once a prompt passes `above_tokens`.\n\nPRICES\n  USD per million tokens, input and output, with cache prices where offered.\n  `blended` is three parts input to one part output. `batch_price` is the rate\n  for work submitted in batches, where the model's maker offers it.\n\nSOURCES\n  The models come from OpenRouter's public listing. The prices come from every\n  place they are published: OpenRouter's hosts, the sellers' own rates (the\n  clouds and the inference hosts), Hugging Face's router, and hosts that sell\n  on no marketplace. Each offer names its `provider`, the route it is sold\n  through (`via`, null when bought direct) and the `sources` that reported it;\n  a source that could not be read is named in `sources_failed`. A price in\n  euros is converted at the day's reference rate and the note says so. Read at\n  most an hour old.\n\nStudies:\n  frontier  \u003cquery\u003e\n      the newest model of every current family, newest first — with its version, release date, context, the models it replaced, and its listed price\n  providers \u003cmodel\u003e\n      every provider serving one model, cheapest first — input, output and cache prices per million tokens, quantization, speed tier, long-prompt price, context and recent uptime — with the cheapest standard offer by input, output and blended, and the cheapest flex\n\nJSON to stdout, so it composes: pipe to `jq` to pick a field, or `\u003e /tmp/…` to keep it.\nRun a study with no arguments — `model-prices frontier` — to see what it takes.\n","example_command":"model-prices providers \"glm 5.3\"","example_output":"","cost_class":"upstream-fetch","tier":"metered","billing_items":["model_prices"],"price_per_1k_calls":3,"price_is_from":false,"status":"available","commands":["model-prices"],"studies":[{"call":"frontier","about":"the newest model of every current family, newest first — with its version, release date, context, the models it replaced, and its listed price","args":[{"name":"query","positional":true,"list":false,"hint":"words the model's id or name must contain — glm, claude, coder — or all"},{"name":"author","positional":false,"list":false,"hint":"only this maker: anthropic, openai, google, z-ai, qwen, deepseek…"},{"name":"max_age_days","positional":false,"list":false,"hint":"leave out families with nothing newer than this; default 365, 0 for no limit"},{"name":"providers","positional":false,"list":false,"hint":"true | false — also read every provider and report each model's cheapest (at most 40 models)"},{"name":"limit","positional":false,"list":false,"hint":"how many models, newest first; default 100, at most 300"}]},{"call":"providers","about":"every provider serving one model, cheapest first — input, output and cache prices per million tokens, quantization, speed tier, long-prompt price, context and recent uptime — with the cheapest standard offer by input, output and blended, and the cheapest flex","args":[{"name":"model","positional":true,"list":false,"hint":"an id (z-ai/glm-5.3) or words (\"glm 5.3\", \"claude opus\"); a family alone means its newest"},{"name":"sort","positional":false,"list":false,"hint":"blended (default) | input | output"}]}],"default_enabled":false,"first_seen":"2026-10-12T07:29:09.42817Z","enabled":false}