Last updated:
Hunyuan
Tencent's Hunyuan family — large open-source mixture-of-experts models for coding, productivity and science.
Model Facts
Latest model: Hunyuan Hy4 Preview
- Provider
- Tencent
- Released
- Aug 28, 2026
- Context window
- 1M+ tokens
- Weights
- Open
Full Model List
| Model | Released | Status | Context | Note |
|---|---|---|---|---|
| Hunyuan Hy4 Preview | Aug 28, 2026 | preview | 1M+ tokens | 770B-parameter MoE (49B active), 1M+ context, Apache 2.0 |
| Hunyuan Hy3 | Jul 6, 2026 | legacy | 256K tokens | 295B-parameter MoE (21B active), 256K context, Apache 2.0 |
- Official model page:
- hunyuan.tencent.com
- Official documentation:
- cloud.tencent.com/document/product/1729
- Other sources:
- Hy4 Preview (LLM Reference)
- Tencent releases (Opper)
What it is
Hunyuan is Tencent's LLM family. Hy4 Preview (August 28, 2026) is a 770B-parameter mixture-of-experts model with 49B active parameters and a context window above 1M tokens, open-sourced under Apache 2.0 and aimed at coding, office productivity, game development and scientific research. Hy3 (July 6, 2026) is a 295B / 21B-active MoE with a 256K native context.
Where this information comes from
Details come from model trackers; Hy4 is labelled a preview, so specifications may change. Release dates, context windows and benchmark scores are vendor-reported unless stated, and harnesses differ between labs, so treat the numbers as indicative.
Notable traits
- Apache 2.0 with no field-of-use or geographic restrictions.
- Very long context.
- Backed by Tencent's consumer and enterprise products.