TIS-100 Bench: Sequence Generator

Turn each input pair into a sequence: the smaller value, the larger value, then a zero terminator.

In TIS-100 Bench, 16 LLMs each had one attempt at this TIS-100 puzzle, with 10 minutes to find a working program and keep improving it, without seeing the community leaderboard. 14 of 16 models solved it. No model found a new Pareto-frontier point beyond the community leaderboard.

14 / 16models solved it
87best LLM cycles (86 community)
0new frontier points

Records

ScoreCommunity bestBest LLMModel
Cycles8687Muse Wonder 1.0 · +1%
Nodes44DeepSeek V4.1 Flash Ties record
Instructions1317DeepSeek V4.1 Flash · +31%

Community records from the community leaderboard when we froze it. A record counts only after fresh-game replay and 100,000 passing random tests.

Pareto frontier

Pareto explorer

Lower is better ↙
Loading interactive Pareto explorer…
Community leaderboardModel frontier pointNew frontier point
Select a point with the mouse or keyboard to inspect its scores and program.

Search over time

Loading puzzle trajectories…

Every model

ModelSolvedBest cyclesBest nodesBest instructionsNew frontier points
01Muse Wonder 1.0Yes87417—
02GPT-6.1 SolYes91418—
03Opus 5.5Yes91418—
04Fable 5.1Yes93417—
05Gemini 4 ArgonYes96417—
06Kimi K3.1 ProYes97417—
07Muse Spark 1.3Yes102418—
08Gemini 3.8 FlashYes106417—
09Sonnet 5.5Yes107417—
10GPT-6 AstraYes110417—
11Haiku 5.5Yes131418—
12Grok 4.7Yes134417—
13DeepSeek V4.1 FlashYes145417—
14GPT-6 LunaYes193428—
15GLM-5.3No————
16MiMo-V2.6-ProNo————

Best scores can come from different programs. New frontier points are not matched or beaten on all three scores by any community program.

Best model programs

These programs are in TIS-100 save format and released under the MIT License. All of this puzzle’s model programs are on GitHub and Hugging Face.

Muse Wonder 1.0 Fewest cycles

87 cycles · 7 nodes · 21 instructions30647.87-7-21.txt on GitHub

Muse Wonder 1.0 program running on 30647 in TIS-100
Node 1
MOV UP, ACC
MOV ACC, RIGHT
MOV ACC, DOWN
Node 2
MOV UP, ACC
MOV ACC, RIGHT
SUB LEFT
MOV ACC, DOWN
Node 3
MOV LEFT, DOWN
Node 5
MOV UP, RIGHT
Node 6
S: MOV UP, ACC
JGZ G
B: MOV RIGHT, DOWN
MOV LEFT, DOWN
T: MOV UP, ACC
JLZ B
G: MOV LEFT, DOWN
MOV RIGHT, DOWN
Node 7
MOV UP, LEFT
Node 9
MOV UP, DOWN
MOV UP, DOWN
MOV 0, DOWN

GPT-6 Astra Fewest nodes

110 cycles · 4 nodes · 20 instructions30647.110-4-20.txt on GitHub

GPT-6 Astra program running on 30647 in TIS-100
Node 1
MOV UP ACC
MOV ACC RIGHT
MOV ACC RIGHT
MOV ACC RIGHT
Node 2
MOV UP ACC
SUB LEFT
JLZ B
A: MOV LEFT DOWN
ADD LEFT
MOV ACC DOWN
MOV UP ACC
SUB LEFT
JGZ A
B: ADD LEFT
MOV ACC DOWN
MOV LEFT DOWN
Node 6
MOV UP DOWN
MOV UP DOWN
MOV 0 DOWN
Node 9
MOV UP DOWN

Sonnet 5.5 Fewest instructions

132 cycles · 5 nodes · 17 instructions30647.132-5-17.txt on GitHub

Sonnet 5.5 program running on 30647 in TIS-100
Node 1
MOV UP, ACC
MOV ACC, DOWN
MOV ACC, DOWN
Node 2
MOV UP, ACC
MOV ACC, DOWN
MOV ACC, DOWN
Node 5
MOV UP, RIGHT
Node 6
MOV LEFT, ACC
SUB UP
JGZ S
MOV LEFT, DOWN
MOV UP, DOWN
JMP E
S: MOV UP, DOWN
MOV LEFT, DOWN
E: MOV 0, DOWN
Node 9
MOV UP, DOWN