Balrog — Epoch AI

Balrog — Epoch AI is run by Epoch AI. It has ranked 41 models, of which the catalogue holds 33, scoring from 0.037 to 0.683.

measured by Epoch AI · the board itself

Place	Model	Metric	Score
1stof 41	GPT 6 Astra	Average progress (max)	0.683
2ndof 41	Claude Opus 5	Average progress (max)	0.634
3rdof 41	GPT-5.6 Sol	Average progress (max)	0.6
4thof 41	gemini-3-pro-preview	Average progress	0.581
5thof 41	Gemini 3.1 Pro Preview	Average progress	0.57
7thof 41	GPT-5.6 Terra	Average progress (max)	0.532
8thof 41	Gemini 3 Flash Preview	Average progress	0.481
9thof 41	GPT-5.6 Luna	Average progress (max)	0.456
11thof 41	Claude Opus 4.5	Average progress	0.435
15thof 41	DeepSeek-R1	Average progress	0.349
16thof 41	Gemini 2.5 Flash	Average progress	0.335
17thof 41	GPT-5	Average progress (minimal)	0.328
18thof 41	Claude 3.5 Sonnet	Average progress	0.326
19thof 41	GPT-4o (2024-05-13)	Average progress	0.323
20thof 41	Claude Haiku 4.5	Average progress	0.312
22ndof 41	Grok 3	Average progress	0.295
23rdof 41	Reka Flash 3	Average progress	0.292
24thof 41	Meta-Llama-3.1-70B-Instruct	Average progress	0.279
25thof 41	Llama-3.2-90B-Vision-Instruct	Average progress	0.273
26thof 41	Llama 3.3 70B Instruct	Average progress	0.23
28thof 41	DeepSeek R1 Distill QWEN 32B	Average progress	0.195
29thof 41	Claude Haiku 3.5	Average progress	0.193
30thof 41	Mistral Nemo	Average progress	0.176
31stof 41	GPT-4o-mini (2024-07-18)	Average progress	0.174
32ndof 41	Llama 3.2 11B Vision Instruct	Average progress	0.168
33rdof 41	Qwen2.5 72B Instruct	Average progress	0.162
34thof 41	Llama 3.1 8B Instruct	Average progress	0.151
36thof 41	Qwen 2 VL 72B Instruct	Average progress	0.128
37thof 41	Phi-4	Average progress	0.116
38thof 41	Llama 3.2 3B Instruct	Average progress	0.101
39thof 41	Qwen2.5 7B Instruct	Average progress	0.078
40thof 41	Llama 3.2 1B Instruct	Average progress	0.066
41stof 41	Qwen2-VL-7B-Instruct	Average progress	0.037

Place

Model

Metric

Score

1stof 41

GPT 6 Astra

Average progress (max)

0.683

2ndof 41

Claude Opus 5

Average progress (max)

0.634