Qwen3.7 Text Embedding Flash

A lightweight multilingual text embedding model trained on Qwen3.7, targeted at cost- and throughput-sensitive scenarios. Compared to text-embedding-v4, it features major upgrades in capabilities such as multilingual coverage (100 to 201 languages), long text processing (32K to 128K), and Sparse Embedding; its overall performance on semantic, cross-lingual, and code retrieval benchmarks remains basically on par, with an improvement of approximately 3% on AIRBench-zh. It supports user-customized embedding dimensions of 256, 512, 768, and 1024, with a default of 1024 dimensions.