Newshot

Tencent Hy4 Officially Open-Sourced, Completing Full Open-Source Release of the Hunyuan LLM Series

AI summary

Tencent's Hunyuan team announced today that the Hy4 large model is officially open-sourced, with model weights released simultaneously on GitHub, Hugging Face, ModelScope, and GitCode for developers to download and deploy, completing the full open-sourcing of the Hunyuan LLM series covering Hy3 preview, Hy3, Hy4 preview, Hy4, and Hy4 flash. Hy4 uses a MoE architecture with 770B total parameters, 49B activated parameters, and a context window exceeding 1 million tokens; the cost-effective Hy4 flash has about 220B total parameters, including roughly 63.5B Engram parameters, activates only about 12B parameters per token, also supports up to a 1-million-token context window, and accepts text and image inputs. Hy4 and Hy4 preview are released under the Apache 2.0 license allowing commercial use and modification, and Tencent Hunyuan also published a Hy4 System Card disclosing evaluation results across dimensions such as cybersecurity and biosecurity.

Translation(model translation, not the source's own text)

Tencent Hy4 Officially Open-Sourced, Hunyuan LLM Series Fully Opened Up

IT Home, November 20 — The Tencent Hunyuan team announced today that the Hy4 large model has been officially open-sourced, with model weights now available simultaneously on four platforms: GitHub, Hugging Face, ModelScope, and GitCode. Developers can download and deploy them on their own.

With this, the Hunyuan LLM series has been fully open-sourced, covering versions including Hy3 preview, Hy3, Hy4 preview, Hy4, and Hy4 flash.

The official Hy4 version was first released on November 4, and what is open-sourced this time is the model weights previously announced. Hy4 adopts a MoE (Mixture of Experts) architecture, with 770B total parameters, 49B activated parameters, and a context window exceeding 1 million Tokens.

As IT Home previously reported, released alongside the flagship Hy4 is the cost-effective Hy4 flash. Hy4 flash has about 220 billion total parameters, including approximately 63.5 billion Engram parameters, activating only about 12 billion parameters per Token. It also supports a context window of up to 1 million Tokens and accepts both text and image inputs. The two form a tiered pairing ranging from flagship capabilities to low-cost, large-scale deployment.

Regarding the open-source license, Hy4 and Hy4 preview use the Apache 2.0 license, permitting commercial use and modification. Tencent Hunyuan also released the Hy4 model safety evaluation card (System Card), disclosing evaluation results across dimensions such as cybersecurity and biosecurity.

According to the official statement, Hy4's training data includes only defensive cybersecurity content such as PoC generation and vulnerability remediation, excluding offensive data such as exploit development, reverse engineering, and penetration. Audits by Tencent Xuanwu and Keen Security Lab found no significant offensive capability risks.

  • [GitHub] https://github.com/Tencent-Hunyuan/Hy4
  • [Hugging Face] https://huggingface.co/tencent/Hy4
  • [ModelScope] https://modelscope.cn/models/Tencent-Hunyuan/Hy4
  • [GitCode] https://ai.gitcode.com/tencent_hunyuan/Hy4
  • [GGUF] https://huggingface.co/AngelSlim/Hy4-GGUF
  • [Open-source license] https://huggingface.co/tencent/Hy4/blob/main/LICENSE
  • [Model safety evaluation card] https://hy.tencent.ai/Hy4_system_card.pdf

Related Articles

  • Tencent Hunyuan Hy4 flash model released: 224B total parameters, 12B activated, cost per task about 5% of Hy4
  • Tencent's executive office calls for re-entrepreneurship in the AI era, official full text released
  • Tencent Hunyuan: Hy4 calls surge, WorkBuddy has scaled up tenfold, Hy3 free access extended
  • Tencent Hunyuan Hy4 official version released, touted as first to rank among the strongest open-source models and built for real productivity
  • Tencent Hunyuan releases Hyra 2.0: building reusable tools for model R&D
  • First introduction of visual understanding, Tencent Hunyuan Hy4 and Hy4 flash AI models coming soon

The translation is machine-generated and individual terms may be off. Nothing is rewritten or cut; the original is authoritative.

Published by IT之家
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗