Launching today
Litmiq
Test your prompts across OpenAI models, side by side
1 follower
Test your prompts across OpenAI models, side by side
1 follower
Compare test cases across OpenAI models, side by side, with per-metric judge scores. Free and browser-only; no signup. Your OpenAI key stays in localStorage. OpenAI bills API usage directly to you at standard rates. Setup autosaves; results export as JSON.
Free
Launch Team
Electros LaunchPad - Hybrid AI EnablerHybrid AI: run it locally, add cloud power only when needed
Promoted
Hi Product Hunt! Litmiq is a browser-based tool for comparing prompt outputs across OpenAI models.
How it works: paste your system prompt, add test cases, select models, and hit Run. Each output gets per-metric scores and a short reason from a judge model. If the judge fails to return a valid score, we say so instead of making one up.
Try it: https://litmiq.com/app
Honest limits: OpenAI models only, you pay OpenAI for API usage with your own key, and LLM-judge scores are estimates. Guides: https://litmiq.com/guides
What would you need before trusting this as a pre-deploy check? Feedback very welcome.