Home/Answers/Model comparison
AI COST LAB / Model comparison

Model comparison Failed? Troubleshooting

When Model comparison fails, troubleshoot in order: file/format, permission/lock, output settings, another tool.

Updated 2026-11-01About 3 minLocal test
Quick answer

When Model comparison fails, troubleshoot in order: file/format, permission/lock, output settings, another tool. Compare models on same task, output length and quality, not price per million alone.

Why it matters

Start with the most common and cheapest causes to narrow it down fast.

Quality differs.

Step by step

  1. Measure input, output and retry tokens per user/task.
  2. Separate cache, batch, embeddings and fine-tuning costs.
  3. Use 30 days of real requests for cost per user, then margin.
  4. Model prices change; baseline from official pricing pages and record the date.

Common mistakes

  • Estimating from input tokens only; output often costs more.
  • Forgetting retries, failures and tool calls.
  • Assuming cache hits by default.
  • Using stale prices for long-term margin.

Checklist

  • Input/output/cache/retry tokens are tracked separately.
  • Batch/async fit and latency requirements are clear.
  • Embeddings/fine-tuning/image/audio are separate.
  • Cost per user and margin use real usage.
  • Price date and fallback model are documented.

Key facts

  • Quality differs.
  • Output length differs.
  • Evaluate on real tasks.

FAQ

How do I troubleshoot a failed Model comparison task?

When Model comparison fails, troubleshoot in order: file/format, permission/lock, output settings, another tool. Compare models on same task, output length and quality, not price per million alone.

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: Nops