AI Model Leaderboard

Name: AllBenched AI model leaderboard
Creator: AllBenched

Agentic work ranking, Model families, sorted by Overall
Rank	Model	Provider	Weights	Overall	Coding	Agents	Reasoning	Math	Knowledge	Cost
63	Qwen3.8 27B	Alibaba	Open	56.8	57.7	56.3	57.0	54.4	52.6	$0.83
100	Qwen3.5 397B-A17B	Alibaba	Open	51.8	–No results	51.7	51.9	52.3	–No results	$1.35
134	Qwen3.6 35B-A3B	Alibaba	Open	45.6	44.1	45.1	49.8	49.9	–No results	$0.36
141	Qwen3 235B-A22B Instruct 2507	Alibaba	Open	44.4	44.8	44.3	44.4	–No results	–No results	$0.31
150	Qwen3 32B	Alibaba	Open	42.1	39.8	42.9	42.5	48.0	–No results	$0.26
156	Qwen3-14B	Alibaba	Open	39.6	37.9	39.7	42.3	47.9	–No results	–No price
159	Qwen3.5 9B	Alibaba	Open	38.8	35.7	38.1	46.5	46.9	–No results	$0.14
165	Qwen3 8B	Alibaba	Open	37.1	27.6	40.3	40.4	46.1	–No results	$0.098
166	Qwen3 235B-A22B Thinking 2507	Alibaba	Open	37.1	47.5	30.6	45.8	48.0	53.4	$0.95
172	Qwen2.5-VL 72B Instruct	Alibaba	Open	34.8	–No results	34.2	37.2	37.0	–No results	$1.01