Our new X account is live! Follow @wizwand_team for updates

Human Preference Ranking on Human Evaluation Elo (test)

1,634Elo Score

(ai, ai+1)

Updated 4d ago

Evaluation Results

Method	Links
(ai, ai+1) 2025.12		1,634	-
(ai, ai+1)ds 2025.12		1,620	-
(a0, a*) 2025.12		1,612	-
(a0, a1) 2025.12		1,525	-
(a, b) 2025.12		1,465	-
base 2025.12		1,383	-
a* SFT 2025.12		1,377	-
GPT-4o-0513 2024.09		1,079	1
Molmo-72B 2024.09		1,077	2
Gemini 1.5 Pro 2024.09		1,074	3
Claude-3.5 Sonnet 2024.09		1,069	4
Llama-3.2V-90B-Instruct 2024.09		1,063	5
Molmo-7B-D 2024.09		1,056	6
Gemini 1.5 Flash 2024.09		1,054	7
LLaVA OneVision-72B 2024.09		1,051	8
Molmo-7B-O 2024.09		1,051	9
GPT-4V 2024.09		1,041	10
Llama-3.2V-11B-Instruct 2024.09		1,040	11
Qwen2-VL-72B 2024.09		1,037	12
MolmoE-1B 2024.09		1,032	13
Qwen2-VL-7B 2024.09		1,025	14
LLaVA OneVision-7B 2024.09		1,024	15
InternVL2-Llama-3-76B 2024.09		1,018	16
Pixtral-12B 2024.09		1,016	17
Claude-3 Haiku 2024.09		999	18
Phi3.5-Vision-4B 2024.09		982	19
xGen-MM-interleave-4B 2024.09		979	20
Claude-3 Opus 2024.09		971	21
LLaVA-1.5-13B 2024.09		960	22
InternVL2-8B 2024.09		953	23
Cambrian-1-34B 2024.09		953	24
Cambrian-1-8B 2024.09		952	25
LLaVA-1.5-7B 2024.09		951	26
PaliGemma-mix-3B 2024.09		937	27