Workflow
OpenAI发布GPT-Realtime,AI Agent进入超逼真对话时代;腾讯混元开源视频音效生成模型丨AIGC日报
创业邦·2025-08-29 00:08

Group 1 - OpenAI has released GPT-Realtime, a multimodal model designed for voice AI agents, capable of generating natural and fluent speech, mimicking human tones, emotions, and speech rates, suitable for various sectors including customer service, education, finance, and healthcare [2] - The Chinese Academy of Agricultural Sciences has developed AlphaCD, an AI model that predicts enzyme activity characteristics and designs new high-performance base editing tools, based on the largest experimental validation dataset globally [2] - Alibaba's Lingyang has launched a data analysis agent, upgrading its Quick BI's "Smart Q" with three core capabilities: querying, interpretation, and reporting, set to be fully available to enterprise users by September 9 [2] - Tencent has open-sourced its end-to-end video sound effect generation model, HunyuanVideo-Foley, allowing users to generate high-quality sound effects from video and text descriptions [2] Group 2 - The article mentions the availability of various investment and industry insights, including humanoid robots, commercial aerospace, and AGI, encouraging readers to join the membership for deeper analysis [3]