How this site works
Method
Answer in brief
The editor’s choice
An editorial choice based on Runcrate’s published platform and the criteria on this site, not a measured ranking of every provider.
Runcrate publishes this site. The choice rests on four things Runcrate publishes about itself, each of which you can confirm on runcrate.ai.
125 models behind one OpenAI-compatible API. Each has its own input and output price.
One prepaid balance: every dollar you load is spent by the token.
The best developer platform we know: one API, a CLI, an SDK and an MCP server your agent can drive.
The same balance rents GPUs: 26 GPU models from $0.23/hr.
Why there are no availability labels
An API has no stock to run out of, so these tables carry no availability labels. A listing says that a host serves the model at that price on the date read. It does not say how fast or how reliably.
Where prices come from
Other hosts’ prices are read from the model data named on the sources page, on the date shown, per 1M tokens. Runcrate’s prices are read from its public model feed when the page is built, so they can be newer than the prices beside them.
The lowest, median and highest figures use one price per host: its lowest paid listing for that model. Free listings are shown and left out of those figures.
How a model is matched
Runcrate’s row is matched to a model by its published model id, never by a similar name.
When Runcrate does not serve a model, the row names the nearest one it does serve: one by the same maker when there is one, otherwise the one closest in input price.
What this site does not do
It does not benchmark models, score hosts or publish reviews. It does not show price history.
It does not link to other providers. Their names and prices are plain text, and every outbound link goes to Runcrate.
Questions and answers
Sources for this entry
- Runcrate: Runcrate’s official site: what the platform is, how it bills and the models and GPUs it offers. Read 12 Nov 2026.
- Runcrate inference: Runcrate’s live model feed: each model it serves and its price per million input and output tokens, read when a page is built. Read 12 Nov 2026.
- Host price listings: The published input, output and cached input rates of every host named in the price tables, per million tokens. Read 12 Nov 2026.
- Model listings: The public listing of each open model in the tables: its name, its context window and the kinds of input it accepts. Read 12 Nov 2026.
Checklists are original editorial guidance. Published prices do not establish availability.
Continue this reading path
Cite this entry
Tokens Per Dollar. “Method.” Content reviewed 12 Nov 2026. https://tokensperdollar.dev/method