Skip to content
Dashboard

What is LLM evaluation? A developer's primer on evals

Copy link to headingWhat an LLM eval actually is

Copy link to headingMost teams are still eyeballing outputs

Copy link to headingThree ways to grade an output, and the order that matters

Copy link to headingHow this runs in CI

Copy link to headingWhat to measure

Copy link to headingGeneric evals manufacture confidence you haven't earned

Copy link to headingWiring evals into your pipeline

Copy link to headingLayer the checks, cheapest first

Copy link to headingGate at three checkpoints

Copy link to headingMock in CI, run real models in staging

Copy link to headingThe AI SDK is already an eval harness

import { generateText, Output } from 'ai';
// Force the judge to return a verdict, not a paragraph of hedging
const { output } = await generateText({
model: yourModel,
output: Output.choice({ options: ['pass', 'fail'] as const }),
prompt: `Does this response correctly answer the question?\n\n${candidate}`,
});
// output is 'pass' | 'fail'

Copy link to headingAgent evals are the next gap to close

Copy link to headingEvals are version control for behavior

Copy link to headingFAQ

Copy link to headingHow many test cases do I need to start?

Copy link to headingHow much do LLM evals cost to run?

Copy link to headingShould I eval in production or only before deployment?

Copy link to headingWhat tools work best with TypeScript and the AI SDK?

Ready to deploy?