Tencent has released WeVisDoc, an end-to-end document parser for page images fine-tuned from Qwen3-VL models. The system converts page images into structured Markdown, including LaTeX formulas and HTML tables.
Tencent has released WeVisDoc-4B, an end-to-end document parser for page images that converts pages into structured Markdown with LaTeX formulas and HTML tables. Fine-tuned from Qwen3-VL-4B-Instruct, the model achieves an Overall score of 95.38 on OmniDocBench v1.6 and ranks first among compared parsers across all four reported settings.
Tencent has released Simple Attention Sparsification (SAS) checkpoints that learn to rank and select KV blocks for each query in Qwen3 models. Unlike methods distilling dense attention scores, SAS adds continuous gates to selected blocks, enabling end-to-end optimization of context ranking with the language-modeling loss.
Tencent has made its AuK 1.5B foundation model for speech generation and editing available as open-source software, including code and model weights.
Tencent has made the AuK foundation model for speech generation and editing available as open source, including code and weights on Hugging Face and ModelScope. The release includes AuK-Flash, a distilled variant designed for fast inference with only 4 steps.
Tencent has released EVIE-4.5B, a high-performance multilingual visual document retrieval model featuring dynamic Prefix-Multivector Retrieval Latency (Prefix-MRL) and training-free Hierarchical Agglomerative Clustering (HAC) token compression.
Tencent has released EVIE-8B, a high-capacity visual document retrieval model featuring 4096-dimensional token representations and bidirectional full attention. The model serves as a teacher foundation for the lightweight EVIE-4.5B Prefix-MRL variant and achieves state-of-the-art performance on visual document retrieval benchmarks.