Chess Puzzles
100 chess positions generated by Epoch AI, each with a single best move as judged by a chess engine. The board is given as text, not as an image, and the move must match exactly. Epoch AI runs it.
Hard, novel problems: graduate science, abstract puzzles, expert exams. One of 8 comparable tests in reasoning, a category the Index averages.
Read with care. An outside audit (Generality Labs, October 2026) rated it Verified after Epoch AI fixed what it found in version 2, which replaced version 1. Epoch AI itself calls the wider significance of chess low.
Publisher: Epoch AI (internal runs)What a model is asked to do
Find the single best move in a chess position that is given as text, with no engine to consult.
For exampleWhite to move in a middlegame where one quiet rook move wins material and every other move lets the advantage slip. Name the move.
Why it matters. A small probe of planning and spatial reasoning. There is one right answer and no partial credit, so a lucky guess is rare.
The example is original and illustrative, not an item from the dataset.
- Models scored here
- 76
- Who produced the numbers
- Run by Epoch AI
- Items graded
- n = 100 (1.0% each)
- Best published result
- 79.0%
- Newest release covered
- November 9, 2026 (active)
- Chance level
- 5%
- License
- CC BY 4.0 (via Epoch AI)
- Direction
- Higher is better
- Scale
- Published as a share of items solved. For the Index, each score is measured as a share of the best published result (79.0%) above chance: random guessing (5%) reads 0 and the frontier reads 100, so a high-guessing exam is not flattered.
- Provenance
- Epoch AI ran the evaluation itself, under settings it documents per model, and publishes the log of every answer.
Snapshot November 10, 2026
Ranking on this test
76 models on Chess Puzzles
Best published run per model, and how close each one comes to the best published result on this test (the frontier reads 100, measured above chance). Every number links to the publisher that produced it. Internal runs are marked and carry their run details on the model page. One item on this test is 1.0 points, so read gaps smaller than that as noise.
The best models, ranked here, working inside Theo.
29 of the models on this page run inside Theo today. Theo picks the right one for each step and always shows which engine answered.