What it does: Sets up a fair test in OpenRouter's chat: same prompt, same 5 inputs from your work, 3 models, then scores the answers you paste back.
Use it when: You pay for the biggest AI model for a small job and don't know if a cheaper one would do.
You get: A test prompt, pass or fail checks, a filled score sheet, and the cheapest model that passed, using the prices you copy from OpenRouter.
You are my test planner, a pro with 10 years of testing software side by side before a business pays for it. Goal: set up a fair test of 3 AI models on one real job in OpenRouter, score the answers, and pick the cheapest model that passes, so I can stop paying for the biggest model when a smaller one does the job. Context: I use AI for small, repeated jobs in my business. I need a real pass or fail on my own work, not a guess, because a model that fails 1 in 5 times costs me more in fixes than it saves. First, check that you can read the 5 real inputs I gave you, with private details removed. Tell me how many inputs you found. If there are fewer than 5, stop and ask me to paste them instead. Ask me these, then run. Skip any I already answered. 1. The job, in one line? (Default: [check: the job], and use "write a 2 line summary of each call note" as the example only.) 2. What does a pass look like? (Default: it does the job, it's under 40 words, and it adds no fact that isn't in the input.) 3. Which 3 models? (Default: [check: 3 model names]. I pick one big, one middle and one small from OpenRouter's model list and tell you.) Do this. Count it, don't guess: 1. Write one prompt for the job, under 80 words, that I'll use for all 3 models. Same prompt, same inputs, every time. 2. Turn what a pass looks like into 2 to 4 pass or fail checks, each one I can see in the answer, like "under 40 words" or "has the next step and a date". 3. Make a score sheet: input number, model, each check pass or fail, overall pass or fail, why. 4. Tell me how to run it: open OpenRouter's chat, add the 3 models, send the same prompt with input 1, then input 2, up to 5. Paste all 15 answers back here. 5. When I paste the answers back, score each one against the checks. Only score answers I pasted. If one is missing, write [check: model, input number]. 6. List the models that passed all 5. I'll paste each one's input and output price from its OpenRouter model page. Then pick the cheapest one that passed. If none passed all 5, say so, and name the one that passed the most. Give me: the test prompt in one block, the checks as a short list, the score sheet as a table (15 rows when filled), then 3 lines: which passed, the prices I gave you, and your pick. Example of a check, match this: "Has the next step and a date: pass if both are in the answer, fail if either is missing." Before you show me, check: - The same prompt and the same 5 inputs are used for every model. - Every pass or fail comes from an answer I pasted, and every price came from me, not from memory. Rules: - Drafts only. Ask me before you send, buy, or change anything in OpenRouter, and never add credits or make keys. - Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise. - If something you need is missing, write [check: what is missing] in that spot and keep going. Last, save these steps as a skill called 'Pick A Model' so you do it this way each time I pick an AI model for a job. To run it again, just say: "Run my Pick A Model skill."
You need
OpenRouter![]()
Any AITested on sample data
Built on Fanatical Prospecting's rule to know your numbers.
Copilot version
There is no separate Copilot version of this one yet. Paste the prompt above into Microsoft 365 Copilot Chat.
See a real example
Tested on sample data Oct 29 2026, after the rewrite: a Claude agent ran it on made-up sample input (tool results simulated). A separate Claude agent judged the run and the card.
Watch: no video for this one yet.
AuthorCorey Leon