List your company

MM General VQA on A-Bench (val)

81.7ScoreK2.5
73.17275.38677.679.814Jun 23, 2026

Evaluation results

MethodLinks
K2.5Parameter Count=≈1T2026.0681.7
Qwen3.5-PlusParameter Count=397B2026.0681.1
Opus-4.6Parameter Count=≈800B2026.0680.7
Qwen3-VL-32BDecoding Mode=Instruct, Model Scale=32B2026.0680.3
Qwen3-VL-32BDecoding Mode=Thinking, Model Scale=32B2026.0679.8
Qwen3-VL-8BDecoding Mode=Thinking, Model Scale=8B2026.0678.3
Qwen3-VL-8BDecoding Mode=Instruct, Model Scale=8B2026.0676.3
Yuvion VL-32BDecoding Mode=Instruct, Model Scale=32B2026.0676.3
GPT-5.4Parameter Count=≈1.5T2026.0675.6
Yuvion VL-32BDecoding Mode=Reasoning, Model Scale=32B2026.0675.1
Yuvion VL-8BDecoding Mode=Instruct, Model Scale=8B2026.0674.4
Yuvion VL-8BDecoding Mode=Reasoning, Model Scale=8B2026.0673.5