Linked and quoted from the labs, the press and the people using them. Times in UTC.
Every development we picked up, including smaller stories beyond the labs. Newest first, times in UTC.
Tencent's Gander processes speech, images, and text while handling background tasks. A "cerebellum" keeps conversation going while a swappable "brain" searches files, writes code, or does other complex work, and users can interrupt or change tasks mid-conversation.
Tencent released WeVisDoc-2B, a Qwen3-VL based vision language model, on Hugging Face under the Apache 2.0 license. The model supports English and Chinese and has an associated arXiv paper.
Tencent published the EVIE-4.5B vision-language model on Hugging Face. The model uses ColBERT late-interaction and multi-vector retrieval for visual document retrieval.
Tencent released Hy4-preview, a text-generation model, on Hugging Face. The listing tags it as a conversational mixture-of-experts model under the Hunyuan family.
Tencent released ContextPilot-8B, an 8B text-generation model built on Qwen3, on Hugging Face. The model targets context management, tool use, and agentic conversational tasks.
Tencent published the WeMM-Embedding-2B model on Hugging Face. It is a multimodal embedding model for text and image embedding, built on qwen3_5.
Tencent's Hy-MT2-7B, a 7B-parameter translation model, is now listed on OpenRouter. It supports 33 language pairs plus five Chinese dialect and minority-language pairs, with structured, delimiter-based, contextual, glossary-based, and style-guided workflows.
Tencent published the AuK model on Hugging Face for text-to-speech, voice cloning, speech generation, editing and enhancement. The model has 3895 downloads and 353 likes.
Tencent released the EVIE-Preview-4.5B model on Hugging Face. It is a vision-language model for visual document retrieval built on the colpali-engine.
Tencent released UI-Mate-27B, an image-text-to-text model, on Hugging Face. It is tagged for computer-use-agent and GUI-agent tasks and is built on qwen3_5.