
腾讯发布 Hy4 flash 模型,内部测试任务成本降至二十分之一
来源标题:Tencent launches Hy4 flash, cutting AI task costs by 95% in internal tests
腾讯推出低成本 AI 模型 Hy4 flash,在内部 WorkIndex 基准测试中,其单任务成本仅为完整版 Hy4 的二十分之一。该模型支持多模态输入和最高 100 万 tokens 上下文窗口,在 85 名用户参与的盲测中用户满意度接近 Claude Haiku 5.5,且成本最低、完成时间最短。模型已上线 WorkBuddy、元宝和腾讯云 TokenHub,WorkBuddy 用户可免费使用至 11 月 30 日。
以下正文同步自 动点科技 TechNode,版权归原站所有,已转换为易读排版。
Tencent on Monday launched Hy4 flash, a lower-cost AI model for everyday work such as research, slide creation and spreadsheet editing. On Tencent’s internal WorkIndex benchmark, Hy4 flash cost one-twentieth as much per task as the full Hy4 model. API list prices are $0.07 per million input tokens, $0.281 per million output tokens and $0.003 per million cached input tokens. Hy4 flash supports multimodal inputs and a context window of up to one million tokens. It has 224 billion parameters and activates 12 billion per forward pass. In a separate blind test in which 85 WorkBuddy users rated outputs across 259 tasks, Tencent said the model scored close to Claude Haiku 5.5 in user satisfaction and had the lowest cost and shortest completion time among the models tested. The model is available in WorkBuddy and Yuanbao and through Tencent Cloud’s TokenHub. WorkBuddy users can access it for free through November 30, and TokenHub is offering API access at half the list price. Tencent plans to release the model’s weights and make it available on OpenRouter. [TechNode reporting]
正文由 探How 从来源站点 RSS 同步,内容未经改写;遇到排版缺失或需要图片、视频时请以原文为准。