Notes

What we saw

A write-up for every model we've run ourselves: how it did at each effort, where it went wrong and what it said along the way. Each one is rewritten when a new result comes in. The numbers are always today's.

  1. Claude Haiku 5.5 · 28 Oct 2026

    Claude Haiku 5.5 gave up on every ladder

    Claude Haiku 5.5 sits 10th on the Index, but all 15 ladders across its five effort settings ended with lives left.

  2. Claude Sonnet 5.5 · 28 Oct 2026

    Claude Sonnet 5.5 scores best below max effort

    Claude Sonnet 5.5 sits 4th on the Index with its high run; at max, all three tests ended on cubes it never answered.

  3. DeepSeek V4.1 Flash · 27 Oct 2026

    DeepSeek V4.1 Flash gave up with lives left

    DeepSeek V4.1 Flash sits 14th on the Index, with three rounds ending in a give-up while lives remained.

  4. GPT-6.1 Terra · 26 Oct 2026

    GPT-6.1 Terra cleared both interactive ladders at max

    GPT-6.1 Terra sits 3rd on the Index, but across seven suites it gave up with lives left 10 times.

  5. GPT-6-Luna · 24 Oct 2026

    GPT-6-Luna gave up with lives left

    GPT-6-Luna sits 17th on the Index, with 13 rounds ending in a give-up before it had exhausted its lives.

  6. GPT-5.6-Luna · 23 Oct 2026

    GPT-5.6-Luna gave up nine times with lives left

    GPT-5.6-Luna sits 12th on the Index, with its best result at max, but giving up with lives left persisted from low through its highest effort.

  7. GPT-6.1-Sol · 21 Oct 2026

    GPT-6.1-Sol did best below max effort

    GPT-6.1-Sol sits 5th on the Index, with a 3×3 one-shot clear at xhigh that none of its four max runs matched.

  8. GPT-6-Astra · 21 Oct 2026

    GPT-6-Astra sits just behind Claude Opus 5.5

    GPT-6-Astra sits 2nd on the Index, reaching the 4×4 Gauntlet at max despite giving up with lives left in four rounds across its efforts.

  9. GPT-5.6-Terra · 21 Oct 2026

    GPT-5.6-Terra gave up with lives left

    GPT-5.6-Terra sits 11th on the Index, with its best result at max, but gave up with lives left in 8 rounds across its five efforts.

  10. GPT-6-Sol · 21 Oct 2026

    GPT-6-Sol gave up with lives left in nine tests

    GPT-6-Sol sits 9th on the Index, with its best result at max, but nine tests at lower efforts ended before it used its remaining lives.

  11. GLM-5.3-Flash · 18 Oct 2026

    GLM-5.3-Flash ran out of output on both 3×3 ladders

    GLM-5.3-Flash sits 13th on the Index, reaching level 4 on all three ladders before output limits ended both 3×3 tests with every life unused.

  12. Gemini 3.8 Flash · 17 Oct 2026

    Gemini 3.8 Flash solved a 4-move scramble in 30 moves

    Gemini 3.8 Flash sits 15th on the Index, with a 30-move solve taking it to level 4 on the interactive 3×3 ladder.

  13. Claude Opus 5.5 · 17 Oct 2026

    Claude Opus 5.5 leads the Index

    Claude Opus 5.5 leads the Index at xhigh and went furthest on both 4×4 Gauntlet ladders, while max ended without answering a 3×3 cube.

  14. Grok 4.5 · 17 Oct 2026

    Grok 4.5 reached level 3 on every ladder

    Grok 4.5 sits 16th on the Index, with a 35-move solution to a cube scrambled in just 3 moves.

  15. Composer 2.5 · 17 Oct 2026

    Composer 2.5 went furthest without feedback

    Composer 2.5 reached level 2 one-shot but stopped at level 1 on both interactive ladders, placing it 18th on the Index.

  16. Grok 4.7 · 17 Oct 2026

    Grok 4.7 solved a 4-move scramble in 41 moves

    Grok 4.7’s best result came at medium, including a 41-move solve; it remains unranked because we have not run its top effort, xhigh.

  17. Gemini 3.1 Pro Preview · 17 Oct 2026

    Gemini 3.1 Pro Preview cleared the 2×2 ladder

    Gemini 3.1 Pro Preview sits 8th on the Index after clearing the 2×2 ladder, but stopped at level 4 on both 3×3 tests.

  18. GPT-5.6-Sol · 17 Oct 2026

    GPT-5.6-Sol cleared the 2×2 but stopped early on 3×3

    GPT-5.6-Sol sits 7th on the Index; at max, it cleared the 2×2 ladder without a miss but gave up on the one-shot 3×3 with 2 lives left.

  19. Claude Opus 5 · 16 Oct 2026

    Claude Opus 5 hit output limits on both 3×3 ladders

    Claude Opus 5 sits 6th on the Index, clearing the 2×2 ladder but ending both 3×3 tests without an answer at level 6.