Research
Research & Benchmarks
Benchmark results and research from the Tiki AI team. When a weakness we fix for a client is common across the industry, the eval behind it hardens into a public benchmark.
What We Publish
More on the way
We're just getting started. New benchmarks and papers will be added as our team publishes them. Here's what to expect.
Benchmarks
Controlled evaluations that show how closing a common weakness produces a repeatable gain.
Research Papers
Peer-reviewed and preprint work from our team and the experts in our network.
Technical Reports
Deep dives into our annotation methodology, quality controls, and evaluation tooling.