Skip to content

Benchmark

LHTB — Terminus-2, 90 min budget

Scores are percentages.

Models measured
23
Scores recorded
23
Sources
1
IQ weight
0.83

Score distribution

36%
Mean
36.7%
Median
11.2%
Std dev
25.5%
10th percentile
49.7%
90th percentile
10.2%
Min
10.2%60.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.83 for chat models, measured over 23 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1DeepSeek V4 Flashterminus2-90min· Harbor Hub Leaderboards60.2%
2Grok 4.5terminus2-90min· Harbor Hub Leaderboards50.5%
3Claude Sonnet 5terminus2-90min· Harbor Hub Leaderboards49.7%
4Claude Opus 4.8terminus2-90min· Harbor Hub Leaderboards49.2%
5Claude Fable 5terminus2-90min· Harbor Hub Leaderboards48.7%
6GPT-5.6 Solterminus2-90min· Harbor Hub Leaderboards45.1%
7GPT 5.5terminus2-90min· Harbor Hub Leaderboards44.5%
8Gemini 3.6 Flashterminus2-90min· Harbor Hub Leaderboards39%
9MiniMax M3terminus2-90min· Harbor Hub Leaderboards38.5%
10Kimi K3terminus2-90min· Harbor Hub Leaderboards37.8%
11Claude Sonnet 4.6terminus2-90min· Harbor Hub Leaderboards37.3%
12Kimi K2.7 Codeterminus2-90min· Harbor Hub Leaderboards36.7%
13GLM 5.2terminus2-90min· Harbor Hub Leaderboards31.6%
14Qwen3.6 Plusterminus2-90min· Harbor Hub Leaderboards31.3%
15DeepSeek V4 Proterminus2-90min· Harbor Hub Leaderboards30.7%
16Qwen3.7 Maxterminus2-90min· Harbor Hub Leaderboards29.6%
17Doubao Seed 2.1 Proterminus2-90min· Harbor Hub Leaderboards28.6%
18Gemini 3.1 Pro Previewterminus2-90min· Harbor Hub Leaderboards27.9%
19GPT 5.4terminus2-90min· Harbor Hub Leaderboards27.2%
20GLM 5.1terminus2-90min· Harbor Hub Leaderboards26.7%
21Kimi K2.6terminus2-90min· Harbor Hub Leaderboards25.5%
22GPT-5.3 Codexterminus2-90min· Harbor Hub Leaderboards21.5%
23Grok 4.20terminus2-90min· Harbor Hub Leaderboards10.2%