DeepSeek-V4.1-Flash launches on B.AI with native multimodal MoE capabilities

B.AI has integrated DeepSeek-V4.1-Flash, a native multimodal MoE model featuring a 552 billion parameter backbone and a Causal Encoder-Decoder architecture. The model supports a 1 million token context window and 384K output, designed for long-cycle programming agents and high-concurrency workflows. Web Chat and API access are now available.