Combined Model Ranking

AI BENCHY Category

See which AI models perform best on Combined, which ones stay reliable, and where the biggest gaps appear. Sort by: Metric ↑.

Models Shown

Average Combined Score

6.3

Best Model

Failure Reasons

With failure reason Wrong answer52 With failure reason Invalid tool call19 With failure reason API error13 With failure reason No answer2 With failure reason Timed out2 With failure reason Did not follow instructions1

Rank	Model	Company	Combined Score	Score	Tests Correct	Response Time (avg)
#8	Claude Opus 4.7 none	Anthropic	9.5	8.9	1/1	18.3s
#55	GLM 5.1 medium	Z.ai	9.5	7.3	1/1	43.1s
#68	Claude Opus 4.8 none	Anthropic	9.5	7.0	1/1	17.7s
#77	Claude Sonnet 4.6 none	Anthropic	9.5	6.8	1/1	23.8s
#113	DeepSeek V4 Pro none	DeepSeek	9.5	5.7	1/1	25.5s
#37	Gemma 4 26B A4B medium	Google	9.6	7.6	1/1	73.5s
#10	Claude Opus 4.8 medium	Anthropic	9.8	8.7	1/1	38.0s
#41	Nemotron 3 Ultra 550b A55b medium	NVIDIA	9.8	7.5	1/1	43.9s
#64	MiMo-V2-Flash medium	Xiaomi	9.8	7.2	1/1	75.7s
#70	GPT-5.4 Nano medium	OpenAI	9.8	7.0	1/1	24.1s
#1	Gemini 3 Flash Preview medium	Google	10.0	9.8	1/1	22.4s
#2	Gemini 3.5 Flash high	Google	10.0	9.6	1/1	22.4s
#3	Gemini 3.5 Flash low	Google	10.0	9.4	1/1	6.44s
#5	Qwen3.7 Max medium	Qwen	10.0	9.1	1/1	19.6s
#6	GPT-5.5 low	OpenAI	10.0	9.0	1/1	9.56s

Combined Ranking