Alephant Cost Control icon

Alephant Cost Control

Alephant Cost Control helps route AI requests through Alephant AI Gateway to track token usage, enforce budget limits, prevent runaway agent spend, and attribute AI costs by key, team, model, provider, or session.

Chat Completion

AI-generated

Summary

Perform a chat completion using Alephant AI gateway, optionally routing through cost control.

Inputs

  • model (required) — Model to use for the chat completion.
  • inputMode — Select input method: either a simple prompt string or a JSON array of message objects.
  • prompt — The prompt text when Input Mode is set to Prompt.
  • messages — JSON array of message objects when Input Mode is set to Messages JSON.
  • temperature — Sampling temperature for the model.
  • maxTokens — Maximum number of tokens to generate.
  • responseFormat — Output format; either plain text or a JSON object.
  • metadata — Optional metadata to attach to the request.
  • additionalOptions — Extra parameters passed directly to the underlying API.

Output shape

Array of normalized chat completion objects

Each item includes fields such as raw, text, usage, model, requestId, finishReason, and optionally workspaceId. Errors are thrown as node operation errors.

Examples

Example 1: Simple prompt completion

model: gpt-4o-mini, inputMode: prompt, prompt: 'Explain token pricing', temperature: 0.7, maxTokens: 150

Example 2: JSON messages input with JSON response

model: gpt-4o-mini, inputMode: messages, messages: '[{"role":"user","content":"Summarize this article"}]', responseFormat: json_object

Example 3: Include metadata and additional options

model: gpt-4o-mini, inputMode: prompt, prompt: 'Generate a checklist', metadata: '{"project":"Q4"}', additionalOptions: '{"store":true}'

Links

Discussion