Can You Tell Which AI Model Made This? — elvex

You could be saving 90% of your token spend by using open-weight models.

GLM Mistral Kimi DeepSeek

But is it worth it? You tell us.

Pick a task, choose your models, and see if the difference in cost is worth it.

Step 1: Pick a task

Each task was run with the same data and prompt across Claude Opus 5, ChatGPT 5.6 Sol, DeepSeek Pro V4, Kimi k3, and Grok

Step 2: Enter your email to reveal the model and get the Practical Work Benchmark Report

Enter your email to see what each model outputs for this task, AND to receive the Practical Work Benchmark Report emailed to you.

This report pulls from 82 sources to break down the intelligence output per dollar spent across the most popular knowledge worker use cases today.
Please enter a valid email address.
No spam. Unsubscribe anytime. We never share your email.
Nothing here yet Pick a task above and the outputs will generate here.

Check your inbox.

Your report is heading to now.