Personal Smoke Test for New Models: "Make a CRM App"
Melvin Vivas · X post · 2026-09-03 · Open on X
Topics: Evaluation (Evals) & Testing, LLM Fundamentals · Level: beginner
Summary
When a new model launches, the creator tries it with one personal smoke-test prompt: "make a crm app". The idea is to keep one repeatable task so you can quickly compare new models against each other.
Key points
- Keep one fixed, repeatable prompt to quickly judge each new model.
- The creator's smoke test is "make a crm app". It is a realistic full-app coding task.
- Running the same task on every release makes models easy to compare side by side.
Try this
- Pick your own standard smoke-test prompt and run it on every new model.
- Build a CRM app as a benchmark task for comparing AI coding models.
More in Evaluation (Evals) & Testing
- Local Evals: Open-Source App for Evaluating Local or OpenAI-Compatible Models
- Building an OCR Evals App with Codex
- Be Skeptical of Model Leaderboards: Muse Spark vs Astra
- Compare Model Outputs with Artificial Analysis MicroEvals
- Testing Gemini 3.8 Flash in Cursor with a CRM Smoke Test
- Cohere Parse: Pricing vs Parse Bench Score