Work With AI Demos
Building an AI product? We work with product companies to understand how their products perform on real-world tasks, investigate limitations, and identify opportunities for improvement.
Tell us about your product and what you'd like to learn. We can explore whether a benchmark, detailed report, or focused research collaboration would be useful.
Explore Our Benchmarks →Tell Us What You're Working On
Share a little about your product and the question you're trying to answer. We'll review your inquiry and get back to you if there's a useful way to collaborate.
What We Can Explore Together
The right research depends on your product, the questions you want answered, and what we've already evaluated.
Benchmark Participation & Feedback
Have a product relevant to one of our benchmarks? Share product information, suggest appropriate configurations, or discuss findings that deserve a closer look.
Detailed Tool Reports
For products with completed, reviewed benchmark research, we can discuss a deeper report on observed behavior, limitations, and opportunities for improvement.
Focused Evaluations & Research
Have a particular workflow, failure mode or evaluation question? Let's discuss whether a scoped research engagement could help.
Research That Remains Useful to Everyone
We welcome input from the companies building AI products. Product feedback can help us understand configurations, limitations, and testing approaches more accurately. Our public Results remain evidence-based: paying for additional research does not buy a favorable verdict.