List your company

Text-Image Machine Translation on MMTIT-Bench other2zh 1.0 (test)

85Gemini-flash Judge ScoreGemini2.5-flash
45.365655.655365.94576.2347Mar 25, 2026

Evaluation results

MethodLinks
Gemini2.5-flashParam.=-, Think=true2026.038576.2680.06
Qwen2.5-VLsft(CPR-Trans)Param.=7B, Think=true2026.0382.8482.2381.04
Qwen3-VL-ThinkingParam.=235B-A22B, Think=true2026.0377.972.1478.81
dots.vlm1Param.=671B-A37B, Think=true2026.0376.7372.1478.45
Seed1.6-ThinkingParam.=-, Think=true2026.0375.8273.978.77
Qwen3-VL-InstructParam.=235B-A22B, Think=false2026.0369.676577.2
MiMo-VLParam.=7B, Think=true2026.0369.5265.2476.15
Seed1.6Param.=-, Think=false2026.0366.3565.4576.55
Dots.ocr+Qwen3Param.=-, Think=false2026.0365.2760.9873.42
Qwen2.5-VLsft(Direct)Param.=7B, Think=false2026.0362.4266.3776.81
MinerU2.5+Qwen3Param.=-, Think=false2026.0349.746.2665.06
Qwen2.5-VL_OriginParam.=7B, Think=false2026.0346.8947.0671.67