Gemini 3.8 Flash Officially Launched on B.AI API, Designed for Long-Cycle Programming and Autonomous Agents
2 hour ago / Read about 0 minute
Author:小编   

On September 4, the B.AI platform announced the integration of the Gemini 3.8 Flash model API, the latest offering from Google DeepMind. As the new flagship model in the Gemini 3 series, it combines cost-effectiveness with low latency, delivering significant performance improvements in software engineering, agent tasks, multi-step reasoning in professional domains, as well as specialized applications in finance, law, and beyond. It supports a 1M Token context window and native multimodal input, optimized for complex reasoning and long-term agent workflows.