B.AI launched Tencent Hunyuan’s Hy4-Flash model on November 17, with access available through both its API and Web Chat. According to Odaily, the multimodal model is designed for everyday office work and frequent productivity tasks, and uses an architecture with 224 billion total parameters, including 64 billion Engram parameters, and 12 billion activated parameters. It accepts text and image inputs, supports a 1 million-token context window, and can be used for deep search, document and chart processing, PowerPoint generation, basic front-end displays, and routine agent workflows. Its decoding speed exceeds 150 tokens per second. B.AI is offering the model at 50% off through November 30, 2026, at 12:00 p.m. SGT.
