< Back to situations

Monitor this situation.

[SITUATION] · [ACTIVE] · [TECHNOLOGY]

4 clusters · 18 sources · 11 days · First seen · Last updated

Tencent Hy4 preview AI model release and deployment

Overview

Tencent has released and open-sourced the Hy4 preview, a next-generation large language model (LLM) utilizing a Mixture-of-Experts (MoE) architecture. The model features 770 billion total parameters, with 49 billion parameters activated per token, and supports a context length of over 1 million tokens.

Designed to enhance productivity in software engineering, scientific research, and office tasks, the model was co-designed with Tencent products such as WorkBuddy and CodeBuddy. In an internal blind evaluation involving 163 experts and 203 engineering tasks, the Hy4 preview achieved an average score of 2.99 out of 4.00, outperforming models including GLM-5.3 and Kimi K3.

Following its integration into the WorkBuddy AI service on August 28, Tencent previously announced an emergency expansion of its inference clusters to address high request volumes and task queues. To encourage adoption, Tencent is offering a free trial period for the Hy4 preview through WorkBuddy and CodeBuddy.

The model is accessible globally via Tencent applications including WorkBuddy, CodeBuddy, Yuanbao, and ima. Developers can also access the model via APIs through Tencent Cloud TokenHub and OpenRouter. In recent benchmarks, the Hy4 preview demonstrated significant improvements over its predecessor, Hy3, ranking eighth globally on the Code Arena WebDev leaderboard.

Entities

Tencent · Hy4 preview · WorkBuddy · Tencent Cloud · Zhipu AI

Timeline

  1. 4 days ago

    [TECHNOLOGY] 2 sources
    Tencent releases open-source Hy4 preview large language model

    Tencent has open-sourced Hy4 preview, a large language model with 770 billion parameters and a 1 million+ token context window, optimized for coding, office tasks, and scientific research.

  2. 10 days ago

    [TECHNOLOGY] 2 sources
    Tencent releases open-source Hy4 preview large language model

    Tencent has open-sourced its Hy4 preview large language model, featuring 770B parameters and advanced coding and productivity capabilities, marking a significant return to the top tier of open-source AI.

  3. 12 days ago

    [TECHNOLOGY] 2 sources
    Tencent expands inference clusters due to Hy4 preview demand

    Tencent is expanding its inference clusters to address high demand and task queues for its new Hy4 preview large language model following its integration into the WorkBuddy AI service.

  4. 15 days ago

    [TECHNOLOGY] 12 sources
    Tencent releases open-source Hy4 preview AI model

    Tencent has open-sourced Hy4 preview, a 770B parameter MoE large language model designed for coding, research, and finance, featuring a 1-million-token context window.

Sources

ad-hoc-news.de · blocktempo.com · brasil247.com · click2macao.com · cnmo.com · gadgetpilipinas.net · ithome.com · kulturegeek.fr · leiphone.com · newsbytesapp.com · scmp.com · slashgear.jp · the-tech-trend.com · timesnownews.com · tmtpost.com · tokenpost.kr · wazzup.ph · webwire.com

This summary has been updated 2 times: see revision history