Overfitting Dicta.™
Legal-AI Benchmark Explorer

Legora BAR Tax Case

Freshworks FY2023 Master File

Independent evaluation by Overfitting DictaOriginal benchmark: GitHub · Apache 2.0 License

Compare AI models across 175 rubric criteria. Explore individual scores, grading rationales, and supporting evidence.

27 × 175 criteria
Benchmark Results12 models · 27 evaluations · 4,725 judgments · 175 criteria

Model Rankings

12 models · 27 evaluations

Ranked by best observed pass rate. Select individual evaluations to compare below.

Select independently scored evaluations
Models ranked by their highest completed evaluation, with their next two evaluations in descending score order. All performance bars use a zero to one hundred percent scale.
Best Observed Pass RateOther Configurations
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position
No completed evaluation recorded in this position

Equal best scores share a rank. Evaluations remain independent; comparison columns stay in selection order.

12 selected

Rubric Comparison

12 selected configurations

Compare selected configurations across individual rubric criteria.

175 of 175 criteria
PASSFAIL·One point per pass
Scores by rubric criterion and evaluated configuration. Overall and section scores retain their full dataset denominators, regardless of search or hidden columns.
Rubric / configuration10 sections · 175 criteriaOpus 5.5Sonnet 5.5Fable 5.1Grok 4.7Muse Wonder 1.0Gemini 4 ArgonKimi K3GPT-6 AstraGPT-6.1 SolMuse Spark 1.3Mistral Large 4Gemini 3.8 Flash
XhighOverall 125 / 175MaxOverall 122 / 175HighOverall 119 / 175XhighOverall 118 / 175HighOverall 114 / 175HighOverall 111 / 175MaxOverall 99 / 175XhighOverall 98 / 175MaxOverall 96 / 175MaxOverall 95 / 175HighOverall 92 / 175HighOverall 84 / 175
Overall pass rateAll 175 rubric criteria · unweighted
125 / 17571.43%
122 / 17569.71%
119 / 17568%
118 / 17567.43%
114 / 17565.14%
111 / 17563.43%
99 / 17556.57%
98 / 17556%
96 / 17554.86%
95 / 17554.29%
92 / 17552.57%
84 / 17548%
5 / 6
6 / 6
6 / 6
6 / 6
5 / 6
6 / 6
4 / 6
5 / 6
6 / 6
4 / 6
4 / 6
4 / 6
27 / 37
27 / 37
24 / 37
18 / 37
18 / 37
20 / 37
20 / 37
13 / 37
13 / 37
13 / 37
16 / 37
12 / 37
13 / 22
13 / 22
14 / 22
14 / 22
14 / 22
15 / 22
12 / 22
11 / 22
12 / 22
17 / 22
13 / 22
12 / 22
17 / 21
19 / 21
16 / 21
16 / 21
17 / 21
15 / 21
15 / 21
14 / 21
16 / 21
14 / 21
14 / 21
12 / 21
20 / 27
21 / 27
21 / 27
22 / 27
19 / 27
17 / 27
15 / 27
17 / 27
14 / 27
14 / 27
15 / 27
15 / 27
9 / 19
7 / 19
9 / 19
10 / 19
9 / 19
9 / 19
8 / 19
8 / 19
7 / 19
7 / 19
7 / 19
7 / 19
9 / 11
7 / 11
8 / 11
10 / 11
8 / 11
8 / 11
6 / 11
7 / 11
7 / 11
6 / 11
6 / 11
8 / 11
5 / 10
3 / 10
3 / 10
3 / 10
4 / 10
4 / 10
2 / 10
3 / 10
2 / 10
2 / 10
2 / 10
1 / 10
18 / 18
17 / 18
16 / 18
17 / 18
18 / 18
15 / 18
16 / 18
18 / 18
18 / 18
17 / 18
14 / 18
12 / 18
2 / 4
2 / 4
2 / 4
2 / 4
2 / 4
2 / 4
1 / 4
2 / 4
1 / 4
1 / 4
1 / 4
1 / 4
Select a criterion or verdict to inspect the work
12 columns

Scoring Rationales

Explore the recorded scoring rationales and supporting evidence for each selected configuration.

Select a criterion in the comparison above to inspect its recorded judgments and evidence.

Scores reflect recorded determinations. Rationales and supporting evidence are preserved as recorded.