Skip to content

Benchmark

Orak 2048

Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages. Higher is better.

Models measured
6
Scores recorded
10
Sources
1
IQ weight
0.62

Score distribution

7.6%
Mean
5.6%
Median
6.4%
Std dev
3.5%
10th percentile
11.5%
90th percentile
1.1%
Min
1.1%25.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.62 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Data credits: Orak leaderboard, KRAFTON AI (Park et al., NeurIPS 2025). Public numeric scores; leaderboard licence not stated (code MIT not applied; CC BY-NC trajectories not used).