Ask a support question

Asks Ana, the Zernio support agent, a question about your workspace. Each question starts a run that answers asynchronously: this returns 202 with a runId, and the answer arrives through the support.run.completed and support.run.failed webhooks. GET /v1/support/runs/{runId} is the fallback. Pass threadId to continue an earlier conversation, and context to point Ana at a post, account or profile. When Ana finds a problem on Zernio's side she opens a support ticket with the team (see supportTicket on the run) and the run ends as needs_human.

Pricing. Each question is billed once, when its run finishes, for the tokens Ana used:

• Input: $6 per million tokens
• Output: $30 per million tokens
• Cache reads: $0.60 per million tokens
• Cache writes: $7.50 per million tokens

Most questions cost between $0.10 and $0.50. A run never bills more than its maxCostUsd ($0.50 to $3, $3 by default): near the cap Ana stops investigating and answers with what she found. Runs that end as needs_human are billed like any other; failed runs are free. The run's usage holds the token counts and costUsd the amount billed. Charges appear as "Support API" on your invoice.

Requires an unrestricted API key, usage-based billing and a card on file. Limits per account: 3 active runs and $100 of runs per UTC month. Send an Idempotency-Key header to make retries safe.

import Zernio from '@zernio/node';const zernio = new Zernio({ apiKey: process.env.ZERNIO_API_KEY });const { data } = await zernio.supportassistant.askSupportQuestion({  body: {    message: 'Hello, world!',  },});console.log(data);

{  "runId": "64f0a1b2c3d4e5f6a7b8c9d1",  "threadId": "64f0a1b2c3d4e5f6a7b8c9d2",  "status": "queued",  "pollAfterSeconds": 5}