Prompt Engineering Sandbox

Test different prompts and models side-by-side. See latency, token usage, and approximate model size — perfect for demoing “why bigger models are slower but better.”

Tips

Change only 1 thing in the prompt at a time.
Compare short vs. long answers.
Show the difference in latency between 3B and 26B.
Keep system prompt consistent across runs.

Model A

AI response

Run a prompt to see output.

Model B

AI response

Run a prompt to see output.