生成式AI 一、 Gemini 2.5 Computer Use发布，让AI直接操作浏览器 1. 谷歌DeepMind发布Gemini 2.5 Computer Use模型，类似OpenAI的CUA，能让AI直接控制用户浏览器执行点 击、滚动和输入等操作； 2. 该模型在相关基准测试中性能达到SOTA水平，使用效率高于竞品，特别在多步骤、长时间、跨标签页任务上表现 突出； 3. Google为该模型内置多层安全机制，包括逐步安全服务和系统指令约束，开发者已可通过Google AI Studio和 Vertex AI的Gemini API获取该能力。 https://mp.weixin.qq.com/s/7j9hC317kcixXz2qiPWVBQ 二、 硬刚Sora 2，马斯克xAI发布视频生成模型Imagine v0.9 1. 马斯克旗下xAI推出视频生成模型Imagine v0.9并向所有用户免费开放，相比初代版本在视觉质量、动作和音频生 成方面有所提升； 2. 该模型视频生成时间不到20秒，支持语音优先界面，能生成6秒左右视频，用户可通过添加自然对话、动态相机效 果等创建电影级效果； 3. 与Sora ...

RT xAI (@xai)Introducing Imagine v0.9, our new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more.Now available for free on all our products: https://t.co/2DPEzEZ03e https://t.co/EzMmKE7V3u ...

Core Insights - The latest video generation model, Imagine v0.9, developed by xAI, has been released for free to all users, potentially as a direct response to OpenAI's Sora 2 model [1][8] - Imagine v0.9 boasts faster video generation times of under 20 seconds, while Sora 2 may take one to two minutes [3] - The model allows users to create videos, images, and text through a voice-first interface, enhancing user experience [1][5] Comparison with Sora 2 - Imagine v0.9 is available for free, while Sora 2 operates on an invitation-only basis [3] - The maximum video length for Imagine v0.9 is approximately 6 seconds, compared to Sora 2's 15 seconds [3] - Despite its advancements, Imagine v0.9 has been noted to have issues with prompt understanding and synchronization between audio and video [3][6] Technical Features - Imagine v0.9 integrates with Grok, allowing for the generation of videos from text or user-uploaded images [5] - Key upgrades include motion control, dynamic camera effects, and the ability to add natural dialogue or expressive singing [5] - The model's custom voice feature raises concerns about deepfake risks, as users can upload images and generate realistic videos of public figures [8] User Experience - Initial user experiences indicate that the web version of Imagine v0.9 is not functioning properly, while the mobile version has connectivity issues [4] - The model does not currently support Chinese language input, which limits its accessibility for non-English speakers [7]

New Grok releasexAI (@xai):Introducing Imagine v0.9, our new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more.Now available for free on all our products: https://t.co/2DPEzEZ03e https://t.co/EzMmKE7V3u ...

Great products coming from @xAI!Ethan He (@EthanHe_42):Excited to share my first project at @xai. Imagine v0.9 is a massive upgrade within just few weeks. No goal is too ambitious for a small team of hardcore engineers. Our model is improving at light speed. Stay tuned for what’s next! ...

Introducing Imagine v0.9, our new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more.Now available for free on all our products: https://t.co/2DPEzEZ03e https://t.co/EzMmKE7V3u ...