Thanks to visit codestin.com
Credit goes to botsbench.com

bots-bench / open weights

How far do open-weight models get?

Only aggregate results are public. Prompts, answers, traces, and run IDs stay private.

Leaders

The best open-weight models, and the closed ones to measure them against

The first three cards are closed models: the highest score, the lowest cost per case, and the strongest proprietary model from China. The open-weight cards follow, best run first.

Chart

Score and cost, every model

Bars show cases solved out of 59. The figure at the right is what the whole run cost. Outlined bars are the GLM 5.3 runs. A dashed outline means Louie.ai reported the run.

Read before comparing

What the bars can and cannot say

    Table

    All open-weight runs

    Click a column heading to re-sort.