Skip to main content

Professional work Text Model Leaderboard

Professional deliverables, business workflows, finance, legal work, and data processing.

Sources
1Sep 30, 2026
100%27 items
72.9
2Sep 28, 2026
100%27 items
67.5
3Sep 1, 2026
88%26 items
67.4
4Sep 22, 2026
100%27 items
66.8
5Nov 9, 2026
75%22 items
66.6
6Jul 24, 2026
88%26 items
64.3
7Oct 17, 2026
88%21 items
63.3
8Jun 9, 2026
75%22 items
63.3
9Sep 2, 2026
88%24 items
62.7
10Aug 12, 2026
88%26 items
62.2
11Sep 21, 2026
100%26 items
61.4
12Sep 2, 2026
100%26 items
60.7
13Sep 3, 2026
100%27 items
60.5
14Sep 21, 2026
88%21 items
60.3
15Aug 3, 2026
75%23 items
59.8
16Jul 9, 2026
88%25 items
59.7
17Oct 5, 2026
100%27 items
59.1
18Oct 8, 2026
100%27 items
58.8
19Jul 31, 2026
50%18 items
58.8
20Sep 22, 2026
75%25 items
58.6
21Jul 9, 2026
75%24 items
58.3
22Sep 20, 2026
75%19 items
58.3
23Apr 23, 2026
75%20 items
58.1
24Jun 30, 2026
75%22 items
58.0
25Aug 5, 2026
63%18 items
58.0
26May 28, 2026
75%22 items
57.9
27Jul 8, 2026
75%20 items
56.8
28Oct 7, 2026
100%27 items
56.3
29Oct 19, 2026
75%15 items
56.0
30Jul 21, 2026
63%20 items
55.9
31Jul 16, 2026
100%26 items
55.9
32Aug 26, 2026
100%24 items
55.8
33Sep 10, 2026
88%25 items
55.0
34Aug 13, 2026
88%23 items
54.7
35Jul 9, 2026
75%25 items
54.4
36Oct 26, 2026
75%18 items
54.3
37Jun 13, 2026
63%20 items
52.9
38Sep 22, 2026
100%27 items
52.4
39Apr 8, 2026
63%16 items
50.0
40Feb 17, 2026
75%19 items
47.6
41Apr 7, 2026
63%17 items
47.0
42May 21, 2026
63%17 items
46.6
43Apr 21, 2026
50%18 items
46.1
44Jun 12, 2026
50%15 items
45.7
45Apr 22, 2026
63%15 items
44.0
46Mar 17, 2026
63%20 items
44.0
47Apr 22, 2026
50%12 items
43.4
48Jun 2, 2026
50%14 items
43.2
49Jul 30, 2026
75%21 items
42.1
50Jun 1, 2026
88%21 items
41.3
51Jul 23, 2026
50%11 items
39.1
52Jul 21, 2026
75%21 items
38.9
53Apr 17, 2026
63%15 items
37.9
54Jul 15, 2026
100%23 items
37.3
55Mar 18, 2026
63%15 items
36.2
56Oct 15, 2025
63%19 items
35.5

Not yet ranked

These models have partial scores and will receive a rank when evidence is complete.

71.4Fewer than 3 coding evaluation types

Frequently asked questions