Presents Hunyuan-A13B, an open-source Mixture-of-Experts LLM with 80B total parameters but only 13B activated per inference, pretrained on a 20T-token corpus with enhanced STEM data curation. It introduces a dual-mode Chain-of-Thought framework that adaptively switches between fast thinking for routine queries and slow thinking for complex multi-step problems. The model achieves competitive results across math, science, coding, and agent tasks relative to much larger models while maintaining high inference throughput for latency-sensitive deployment.
