Tencent Open-Sources 770B MoE Hunyuan Model, Hits #5 on Arena Coding
机器之心 · wechat · 2026-08-28
Tencent Hunyuan has open-sourced Hy4preview, a 770B-parameter MoE model with 49B active parameters per inference, 1M context support, and 31.8% higher end-to-end throughput over baseline, with the full training and optimization pipeline included.
It targets practical use: ranked #5 in Arena coding tests (#3 among open models) and beat GLM and Kimi in internal blind tests across 200+ engineering tasks, focusing on coding, office work, game dev, and research. Weights and code are fully open with a permissive commercial license; API pricing is ¥6 per million input tokens and ¥18 per million output tokens, with two weeks of free access via WorkBuddy and CodeBuddy.
Related event: Tencent Open-Sources Hunyuan Hy4 Preview: 770B MoE with 1M Context(28 posts)→
More from coding & agent
- Vercel Goes Full Backend: Python GA, Containers, and Daemons — cramforce · 2026-10-16
- WorldMemArena Accepted at NeurIPS: Memory Storage ≠ Memory Use — xwang_lk · 2026-10-16
- Vercel Ship Demos Container Deploys for Go, Ruby, Java and More — edwin · 2026-10-16
- TRL adds day-0 support for fine-tuning Muse Spark Open's 816B MoE with LoRA on 64 H100s — QGallouedec · 2026-10-16
- Databricks engineers land impactful PRs on day 2 with AI dev tools, vs week 3 before — Yuchenj_UW · 2026-10-16
- Perplexity open-sources SPACE, the sandbox platform powering its Computer agents — AravSrinivas · 2026-10-16