Usage guide
Setup, environment variables, and Python/JavaScript/Go usage examples.
CerebrasLLM
Text generation on Cerebras inference hardware, using your own Cerebras API key. Choose it for very fast token output with open-weight models. Authenticates with CEREBRAS_API_KEY from your environment, or an explicitapi_key.
Fields
str
default:"UNSET"
Cerebras model name, e.g. “gpt-oss-120b” or “qwen-3.8-27b”. Defaults to ‘gpt-oss-120b’.
float
default:"UNSET"
Sampling temperature, 0 to 2. Lower values give more focused, repeatable replies. Defaults to 0.7.
Literal['auto', 'required', 'none']
default:"UNSET"
How the model may call tools: “auto” lets it decide, “required” forces a tool call, “none” disables tools. Defaults to ‘auto’.
int | None
default:"UNSET"
Maximum number of tokens the model may generate per reply. Leave unset for no explicit limit. Defaults to None (not set).
float | None
default:"UNSET"
Nucleus sampling, 0 to 1: only the most likely tokens up to this cumulative probability are considered. Defaults to None (not set).
int | None
default:"UNSET"
Random seed for more reproducible outputs across identical requests. Defaults to None (not set).
str | None
default:"UNSET"
Text sequence at which the model stops generating. Defaults to None (not set).
str | None
default:"UNSET"
Identifier for your end user, useful for usage tracking and abuse monitoring. Defaults to None (not set).
str | None
default:"UNSET"
How much a reasoning model thinks before answering: “low”, “medium” or “high”. Only applies to reasoning models. Defaults to None (not set).