Deepswe — Epoch AI

Deepswe — Epoch AI is run by Epoch AI. It has ranked 69 models, of which the catalogue holds 26, scoring from 0.117 to 0.741.

measured by Epoch AI · the board itself

Place	Model	Metric	Score
1stof 69	GPT 6 Astra	Pass@1 (xhigh)	0.741
2ndof 69	Gemini 3.8 Flash	Pass@1 (high)	0.738
3rdof 69	Claude Opus 5	Pass@1 (max)	0.736
9thof 69	GPT-5.6 Sol	Pass@1 (max)	0.727
12thof 69	Claude Fable 5	Pass@1 (xhigh)	0.699
14thof 69	GPT-5.6 Terra	Pass@1 (max)	0.696
16thof 69	GLM-5.3	Pass@1 (max)	0.69
19thof 69	Kimi K3	Pass@1 (max)	0.685
20thof 69	Grok 4.6	Pass@1 (medium)	0.675
21stof 69	GPT-5.6 Luna	Pass@1 (max)	0.672
22ndof 69	GPT-5.5	Pass@1 (xhigh)	0.67
25thof 69	Gemini 3.7 Flash	Pass@1 (medium)	0.655
30thof 69	GLM 5.3 Flash	Pass@1 (max)	0.634
34thof 69	Claude Opus 4.8	Pass@1 (max)	0.59
36thof 69	Qwen 3.8 Max	Pass@1 (xhigh)	0.575
38thof 69	Muse Spark 1.2	Pass@1 (xhigh)	0.549
41stof 69	Claude Sonnet 5	Pass@1 (max)	0.538
43rdof 69	Grok 4.5	Pass@1 (high)	0.538
45thof 69	Muse Spark 1.1	Pass@1	0.533
47thof 69	GPT-5.4	Pass@1 (xhigh)	0.518
51stof 69	Gemini 3.6 Flash	Pass@1 (high)	0.467
54thof 69	GLM 5.2	Pass@1 (max)	0.438
58thof 69	Gemini 3.5 Flash	Pass@1 (medium)	0.374
62ndof 69	Kimi K2.7 Code	Pass@1	0.305
64thof 69	Claude Sonnet 4.6	Pass@1 (high)	0.299
67thof 69	Gemini 3.1 Pro Preview	Pass@1	0.117

Place

Model

Metric

Score

1stof 69

Pass@1 (xhigh)

0.741

2ndof 69

Pass@1 (high)

0.738