Skip to content

Benchmark

Vals Tax Agent Bench

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

Models measured
85
Scores recorded
85
Sources
1
IQ weight
2.00

Score distribution

25.1%
Mean
25.8%
Median
12.8%
Std dev
7%
10th percentile
43%
90th percentile
2.5%
Min
2.5%52.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.00 for chat models, measured over 85 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Showing 51–85 of 85 results.