Home/Answers/Evaluation
TOKEN LAB / Evaluation

How to Choose Evaluation

Choose Evaluation by use case and compatibility first, specs second; test locally when possible.

Updated 2026-10-31About 3 minLocal test
Quick answer

Choose Evaluation by use case and compatibility first, specs second; test locally when possible. Token budgets affect eval size/cost; sample first.

Why it matters

Stacked specs do not guarantee fit for your use case.

Long eval sets cost more.

Step by step

  1. Confirm the model context and output budget.
  2. Paste the prompt/document and estimate tokens/usage.
  3. Above 80-90%, split, summarize or retrieve.
  4. Count system prompts/tool schemas too.
  5. Verify billing/hard limits with the official tokenizer.

Common mistakes

  • Treating character count as token count.
  • Ignoring output/system prompt usage.
  • Assuming all models use the same tokenizer.
  • Stuffing a long document instead of retrieving/summarizing.

Checklist

  • Model context confirmed.
  • Output budget reserved.
  • System/tool prompts counted.
  • Long text retrieved/split.
  • Critical cases checked with the official tokenizer.

Key facts

  • Long eval sets cost more.
  • Stratified sampling covers cases.

FAQ

How do I choose Evaluation?

Choose Evaluation by use case and compatibility first, specs second; test locally when possible. Token budgets affect eval size/cost; sample first.

How long does it take?

Usually 1-3 minutes; for intermittent issues, run 30 seconds and retest elsewhere.

Is data uploaded?

Runs locally and does not upload video/audio; extensions may still read the page.

Source: OpenAI evals