Baidu released ERNIE 5.5 Flash Base

What happened: Baidu released ERNIE-5.5-Flash-Base, a sparse Mixture-of-Experts language model, as an open base checkpoint at the end of pre-training, before any mid- or post-training. Key numbers: 280B total parameters, 14B activated per token; pre-trained on 33.5T tokens; training context up to Why it matters: The release is positioned as a starting point for domain adaptation, instruction tuning and reinforcement-learning research, intended to spare developers the cost of large-scale pre-training. Efficiency was a stated design goal across attention, MoE layers and long-context cost.

Baidu released ERNIE-5.5-Flash-Base, a sparse Mixture-of-Experts language model, as an open base checkpoint at the end of pre-training, before any mid- or post-training.

The release is positioned as a starting point for domain adaptation, instruction tuning and reinforcement-learning research, intended to spare developers the cost of large-scale pre-training. Efficiency was a stated design goal across attention, MoE layers and long-context cost.