Compare Chat Models
A tool for comparing chat models with your own eyes. Pick 2-4 models, give one prompt, and the system runs them together and puts the results next to each other with the time and cost each one took.
What you provide: text · What you get back: chat
Why compare for yourself
Published benchmark scores describe the general case, not whether a model suits your work. Running the prompts you actually use every day tells you far more than a scoreboard.
- Find the best value model for work you repeat daily
- Check whether a job really needs the expensive model
- Compare how each one writes and structures an answer
- See whether a cheap model is good enough for simple tasks
What the page shows
- The full answer from every model, side by side
- How long each one took
- The actual cost per model
- Web sources used, where a model searched
About the cost
Each model is billed separately — choosing four means paying four times. It is a one-off cost to learn which model to use from then on, which usually beats overpaying for an unnecessarily expensive model every day.
Frequently asked questions
How many models at once?
Between 2 and 4 per comparison.
Can I see past comparisons?
Yes, saved comparison runs can be reopened without re-running them and without paying again.