Qwen-MT-Image (Image Translation)
'/%3e%3c/svg%3e "Model supports image analysis")
API Overview
Qwen-MT-Image, the Tongyi Qianwen image translation model, supports translating text in images from 11 languages into either Chinese or English. Leveraging advanced OCR and NMT technologies and trained on an ultra-large-scale dataset of text-image pairs, it can accurately recognize various fonts—including handwritten and printed styles—as well as complex layouts featuring both horizontal and vertical text arrangements. During translation, it precisely preserves the original content and layout style. Additionally, it offers customizable features such as term definitions to ensure accurate translations in specialized fields, sensitive word filtering to maintain content compliance, and image subject detection to safeguard the visual integrity of key areas. This makes it an efficient and high-quality solution for scenarios like cross-border e-commerce and multilingual document processing.
API Console
Log in to explore more features! Click to Log In