Kayts gives you five models in one chat, from Kayts Mihr, fast and free, to Kayts Monte, our most powerful. Each one runs on a model built by another company, chosen by us for that job. Those makers test their models on public benchmarks and publish the scores. We copy those scores as they were printed; we did not run the tests ourselves.
One test, three models
Only three of the five makers report the same test, GPQA Diamond: 198 hard multiple-choice questions in biology, chemistry and physics, written so that experts in the field get most right and skilled non-experts with web access do not. It is one of the few places where the scores line up directly.
Model cards of the models behind Kayts Mihr, Anahit and Nareia, checked 20 October 2026.
Show the data as a table
| Value | |
|---|---|
| Kayts Mihr | 81.7% |
| Kayts Anahit | 89.2% |
| Kayts Nareia | 90.1% |
All three answer most of these questions correctly; the spread is 8.4 points. Kayts Mihr, the one in every plan, sits closest to the bottom, and Kayts Nareia, from the Pro plan, at the top.
Every model, on the tests its maker chose
Most makers pick the tests that fit what their model is for: coding and terminal work for one, long documents for another. So each panel below has its own set, and a bar in one panel should not be read against a bar in another, even when the names look alike. Versions matter too: Terminal-Bench 2.1 and 4.0 are different tests, and Humanity's Last Exam with tools is far easier than without.
Kayts Mihr
Kayts Anahit
Kayts Siran
Kayts Nareia
Kayts Monte
Sources: the model cards and announcements of the models behind each Kayts model, checked 20–21 October 2026 (listed in the Privacy Policy).
Show the data as a table
| Model | Benchmark | Score |
|---|---|---|
| Kayts Mihr | GPQA Diamond | 81.7% |
| Kayts Mihr | MMLU-Pro | 82.5% |
| Kayts Mihr | HMMT February 2025 | 83.2% |
| Kayts Mihr | LiveCodeBench v6 | 65.6% |
| Kayts Mihr | τ²-Bench | 79.1% |
| Kayts Mihr | MMMU | 78.4% |
| Kayts Anahit | GPQA Diamond | 89.2% |
| Kayts Anahit | Humanity's Last Exam | 30.8% |
| Kayts Anahit | LiveCodeBench v6 | 90.3% |
| Kayts Anahit | SWE-bench Pro | 61.7% |
| Kayts Anahit | Terminal-Bench 2.1 | 73.0% |
| Kayts Anahit | OSWorld-Verified | 84.3% |
| Kayts Siran | Terminal-Bench 4.0 | 42.3% |
| Kayts Siran | DeepSWE v1.1 | 77.0% |
| Kayts Siran | AutomationBench v1.0.6 | 60.5% |
| Kayts Siran | Humanity's Last Exam | 42.6% |
| Kayts Siran | Agents' Last Exam | 35.4% |
| Kayts Siran | Chartography, with tools | 83.0% |
| Kayts Nareia | GPQA Diamond | 90.1% |
| Kayts Nareia | Humanity's Last Exam | 37.7% |
| Kayts Nareia | LiveCodeBench | 93.5% |
| Kayts Nareia | SWE-bench Verified | 80.6% |
| Kayts Nareia | BrowseComp | 83.4% |
| Kayts Nareia | MRCR 1M | 83.5% |
| Kayts Monte | Terminal-Bench 4.0 | 66.4% |
| Kayts Monte | FrontierCode v1.1 | 54.4% |
| Kayts Monte | CursorBench 4.0 | 57.8% |
| Kayts Monte | Humanity's Last Exam, with tools | 67.7% |
| Kayts Monte | Terminal-Bench-Science 0.1 | 58.7% |
| Kayts Monte | OSWorld 2.1 | 81.8% |
How to read these numbers
- They are the makers' numbers. Each comes from the maker's own model card or announcement, under the conditions it describes (thinking on, the highest effort, tools or none). Which company makes which model, and the links to the sources, are in our Privacy Policy.
- Benchmarks are not your work. A score says how a model does on a fixed set of questions. It does not say how well it writes your email, reads your spreadsheet or understands your field. Try the models on your own tasks; the chat lets you switch between them in one conversation.
- They change. When we move a Kayts model to a newer model, its scores here change with it. Each model's page shows the current set.
More about each model, what it is best at and which plan includes it: all models.
