---
format: "aidr-story-markdown/v1"
id: "c7a81ceee4259ffcdb2cad6349590b76544d2dd12ae7e334e48232cdaba01e58"
canonical_url: "https://aidr.today/c7a81cee?lang=en"
title: "Tencent Opens 770B Hy4 Model to Rank No 5 in Code Arena WebDev"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-08-28T08:42:19.000Z"
category: "Models"
topics: ["tencent","open-source","coding","llm","agent","benchmark"]
source_urls: ["https://x.com/TencentHunyuan/status/2093222928720761009","https://x.com/xeophon/status/2093219350903706070","https://x.com/sgl_project/status/2093224690387165468","https://x.com/arena/status/2093224696745492802","https://x.com/kimmonismus/status/2093237109708468361","https://x.com/testingcatalog/status/2093229582992154830","https://x.com/vllm_project/status/2093248073057357905"]
summary: "The Mixture-of-Experts system features 49 billion active parameters and a 1 million token context window for productivity and office workflows. Tencent released the Hy4 Preview under an Apache 2.0 license to compete in the open source frontier of coding and scientific research. The model scored 1,633 points in the Code Arena WebDev AutoEval, ranking approximately 5th overall and 3rd among open weights models. This performance exceeds the previous Hy3 version, which ranked 31st overall, and includes agentic capabilities that coordinate multiple Codex sessions in parallel to adjust research directions."
---

# Tencent Opens 770B Hy4 Model to Rank No 5 in Code Arena WebDev

> [Open the canonical story](<https://aidr.today/c7a81cee?lang=en>)

**Published:** 2026-08-28T08:42:19.000Z
**Category:** Models
**Topics:** tencent, open\-source, coding, llm, agent, benchmark

## Summary

The Mixture\-of\-Experts system features 49 billion active parameters and a 1 million token context window for productivity and office workflows\. Tencent released the Hy4 Preview under an Apache 2\.0 license to compete in the open source frontier of coding and scientific research\. The model scored 1,633 points in the Code Arena WebDev AutoEval, ranking approximately 5th overall and 3rd among open weights models\. This performance exceeds the previous Hy3 version, which ranked 31st overall, and includes agentic capabilities that coordinate multiple Codex sessions in parallel to adjust research directions\.

## Sources

- [Story source](<https://x.com/TencentHunyuan/status/2093222928720761009>)
- [Supporting source](<https://x.com/xeophon/status/2093219350903706070>)
- [Supporting source](<https://x.com/sgl_project/status/2093224690387165468>)
- [Supporting source](<https://x.com/arena/status/2093224696745492802>)
- [Supporting source](<https://x.com/kimmonismus/status/2093237109708468361>)
- [Supporting source](<https://x.com/testingcatalog/status/2093229582992154830>)
- [Supporting source](<https://x.com/vllm_project/status/2093248073057357905>)