Hunyuan-A13B

model paper
Tencent
High-efficiency open-source MoE model (80B total / 13B active) with dual-mode reasoning and 256K context. Frequently used in benchmarks and serves as the backbone for HunyuanImage 3.0.
HuggingFaceGitHubTechnical Report (PDF)Technical Report (arXiv)Announcement

Outputs 2

Hunyuan-A13B

model
HuggingFaceGitHub
Architecture MOE
Parameters 80B
Active params 13B
Context window 256,000

Hunyuan-A13B Technical Report

paper

Technical report for the 80B total / 13B active MoE model with dual-mode reasoning capability, pretrained on a filtered 20T-token corpus. First published as a PDF in the GitHub repository at release; posted to arXiv on 2026-09-23.

Technical Report (PDF)Technical Report (arXiv)
arXiv HTML
moeopen-weightefficiency