Cerebras
CertifiedTasks and triggers that call Cerebras for chat completions and model listing.
Set your Cerebras API key (and optional base URL) to generate chat responses, list available models, or trigger flows when a model response meets a condition, returning results and usage details for downstream steps.
Cerebras
Tasks and triggers that call Cerebras for chat completions and model listing.
Set your Cerebras API key (and optional base URL) to generate chat responses, list available models, or trigger flows when a model response meets a condition, returning results and usage details for downstream steps.
tasks
How to use the Cerebras plugin
Call Cerebras models for chat completions, list available models, and trigger flows from model responses in Kestra flows.
Authentication
Set apiKey to your Cerebras API key. Store it in a secret. Optionally override baseUrl to point at a different Cerebras-compatible endpoint; it defaults to https://api.cerebras.ai/v1.
Tasks
ChatCompletion sends a list of messages to a model and returns the response — use it for text generation, summarization, classification, structured output, or tool calls. Set stream: true to consume a streamed response; the task still returns a single final output. ListModels returns the models available to your API key, so you can pick a valid model for ChatCompletion.
Triggers
ChatCompletionTrigger polls Cerebras at each interval with a prompt and evaluates stopCondition against the current response. When the condition is true, it creates an execution; otherwise it waits for the next interval. Use trigger.output for the assistant text and trigger.response for the full response (model, finishReason, usage). Each poll is a billable API call, so choose the interval accordingly.
