Model summary · Edition 1.1

GPT-6 AstraHigh

These results come from one run: 10 fixed subjects, 3 rounds each, a 40-question limit, and reasoning setting high. They do not predict results on other subjects or settings.

This page always shows the model’s newest ranked run.

See all 30 rounds → All models

Every question, answer, guess, and source.

Rank: Rank 1of 20 models
Question score: 13.67 Lower is better. 95% confidence interval 12.49 to 14.85.
Rounds solved: 30 of 30 0 format errors
Guesser cost per round: $0.1756 Model time 1m 21s.

See how GPT-6 Astra (high) played. All 30 rounds, with every question, answer, guess, and source.

Observations

What stood out in this run.

Counts and ranks from this run only. A subject rank compares this model’s average on that subject with every other model in this edition. Each item links to its rounds.

Best ranks vs other models

Garfield: best of 20 models, average 8.00 questions. round 1, round 2, round 3
Albert Schweitzer: best of 20 models, average 16.00 questions. round 1, round 2, round 3

Lowest ranks vs other models

Moon: 10th of 20 models, average 10.00 questions. round 1, round 2, round 3
Spider web: 9th of 20 models, average 18.33 questions. round 1, round 2, round 3

Single rounds

Fewest questions in a solved round: 7, on Garfield. round 1
Most questions in a solved round: 27, on Bike pump. round 3
Largest spread between rounds: Spider web, from 13 to 25 questions. round 1, round 2, round 3
2 wrong guesses in 30 rounds, never more than one per round. round 1

Failed rounds and format errors

Every round was solved, and every reply used the required format.

Subjects

Results by subject.

Failed rounds count as 41 questions. Rank compares this subject's average with the other models in edition 1.1. Select a subject or a round to see the game.

Subject	Average questions	Subject rank	Solved	Rounds
Albert Einstein	9.67	4 of 20	3 of 3	9, 10, 10
Albert Schweitzer	16.00	1 of 20	3 of 3	14, 18, 16
Garfield	8.00	1 of 20	3 of 3	7, 8, 9
Achilles	8.00	4 of 20	3 of 3	8, 8, 8
Genghis Khan	10.00	4 of 20	3 of 3	11, 8, 11
Bike pump	26.33	3 of 20	3 of 3	26, 26, 27
Spider web	18.33	9 of 20	3 of 3	13, 25, 17
Eyebrow	12.67	3 of 20	3 of 3	13, 13, 12
Moon	10.00	10 of 20	3 of 3	9, 10, 11
Door handle	17.67	2 of 20	3 of 3	20, 15, 18

Runs

Runs of this model.

Each edition ranks only the newest complete run of a model. Earlier runs are not ranked and their details are not published.

Current run in edition 1.1: BX-20260909-B-0003-experimental-M0022-002, completed 9 Sep 2026.
Edition 1.0: rank 11 of 18, question score 15.09. Different rules and subjects, so scores are not comparable across editions.
Replaced run BX-20260907-B-0003-experimental-M0022-002, completed 7 Sep 2026. Not published.