The sandbox for every
AI search hypothesis
Your core report gives you the weekly baseline. Prompt Lab is where you test the extra ideas: new pages, launch copy, competitor angles, local prompts, and anything your team wants to validate before the next report.
Demo data for illustration.
When the report is not enough
open the lab
Prompt Lab is the place for experiments, not another static dashboard. If a sales call, launch plan, city page, or competitor claim raises a question, you can test it immediately and decide whether it deserves monitoring.
Core report
100Stable measurement across the prompts Rankry builds for your market.
Prompt Lab
DailyCustom experiments for every question you want to test outside the core report.
Monitoring
24/7Promote a prompt into daily or weekly tracking and let Rankry keep it alive.
Build prompts like assets
not one-off guesses
Use variables, suggestions, model selection, clone, edit, and budget checks to turn rough ideas into repeatable AI-search tests.
Variables
Drop in {brand}, {city}, {country}, {niche}, and more from your project profile.
Suggestions
Start from Rankry-generated prompt ideas when you do not want a blank textarea.
Clone & edit
Duplicate a working prompt, change the angle, switch models, and re-run cleanly.
Request control
Every selected model and monitoring cadence affects how many requests the test will reserve.
Platform for collaborative wiki and project tracking
ChatGPTGeminiPerplexityGrokClaudeDemo data for illustration.
Every model gets
its own evidence
A prompt result is not one averaged number. You see mention status, position, sentiment, response excerpts, and cited sources model by model, so weak spots are obvious.
53%#4.27122m agoAI-powered productivity tool with content generation
ChatGPT✓ Yes#2PositiveClaude✓ Yes#1PositiveGemini✓ Yes#8PositiveNotion AI is integrated directly into the Notion workspace, making it ideal for content generation alongside task and project management. It assists with writing, summarizing, organizing, and streamlining content creation.
PerplexityNo--Grok✓ Yes#3PositiveDemo data for illustration.
You control the sandbox
from idea to evidence
Prompt Lab gives power users room to work. Create prompts, organize them, clone what works, archive what is done, and spend request budget where the next insight matters.
Search
Find any prompt by phrase, city, competitor, model, or outcome.
Clone
Duplicate a useful test and change one variable instead of starting over.
Edit
Change text, models, frequency, and let Rankry reset history when it should.
Archive
Clean the active list without losing old experiments.
Budget
Spend requests intentionally across manual runs and monitoring holds.
Sources
Open the response and citations behind every model-level result.
Spend requests
like experiment capital
One prompt across five models costs five requests. Daily monitoring reserves the future runs before they happen, so the lab stays flexible without surprise overages.
- Run cost is visible before you save.
- Monitoring hold shows the future commitment.
- Available requests update as experiments run.
Demo data for illustration.
Quick answers
How is Prompt Lab different from the core report?
What should I test first?
What counts as a request?
When should I turn on monitoring?
What do I get after a run finishes?
Every hypothesis deserves
a clean AI test
Open the lab, build the prompt, pick the models, spend the requests, and monitor the results that matter between reports.
Test a Custom Prompt7-day free trial / no credit card / first prompt running in minutes
ChatGPTClaudeGeminiPerplexityGrok