gemini-3.5-flash-low · default

{"zh": "Gemini 3.5 Flash 的低思考强度档(`-low` 后缀):响应更快、输出 token 更少,适合对延迟和成本敏感、推理要求不高的场景。模型本身见 gemini-3.5-flash(原生多模态、100 万 token 上下文)。", "en": "The low-thinking variant of Gemini 3.5 Flash (`-low` suffix): faster responses and fewer output tokens, for latency- and cost-sensitive work that doesn't need deep reasoning. See gemini-3.5-flash for the model itself (natively multimodal, 1M-token context)."}