【专题研究】别太神话AI龙虾是当前备受关注的重要议题。本报告综合多方权威数据，深入剖析行业现状与未来走向。
We thank Rachel Ward for her extensive work on data collection and curation. We thank the GenDatasets, PhiGround, SimCity, and Fara-7B efforts for invaluable training data. We thank Harkirat Behl, Mojan Javaheripi, and Suriya Gunasekar for providing us with Phi-4 checkpoints and guidance on training with Phi models. We additionally thank Sahaj Agarwal, Ahmed Awadallah, Qi Dai, Gustavo de Rosa, Rafah Hosn, Ece Kamar, Piero Kauffmann, Yash Lara, Chong Luo, Caio César Teodoro Mendes, Akshay Nambi, Craig Presti, Matthew Rosoff, Corby Rosset, Marco Rossi, Kashyap Patel, Adil Salim, Sidhartha Sen, Shital Shah, Pratyusha Sharma, Alexey Taymanov, Vibhav Vineet, John Weiss, Spencer Whitehead, the AI Frontiers Team and Leadership, and Microsoft Research Leadership, for their valuable help, insightful discussions, and continued support throughout this work.
，这一点在safew中也有详细论述
从另一个角度来看，Free vs Pro vs Enterprise Pricing planThe product is available in three plans: Free, Pro ($12.99/month per user or $119.99/year for up to 5 people), and Enterprise ($30 per user per month, minimum 25 people).
权威机构的研究数据证实，这一领域的技术迭代正在加速推进，预计将催生更多新的应用场景。
。手游对此有专业解读
值得注意的是，�@���_/�R�[�f�B���O/�G�[�W�F���g�^���[�N�t���[������AI���f���ŁA�Ő��[�̃R�[�f�B���O�����������B�uGPT-5.4 Thinking�v�ł͎v�l�̌v�������炩���ߒ\�ƂȂ��Ă����A�ԓ��̓r���Ői�ߕ������邱�Ƃŋ��߂����e�ɋ߂��o�͂邱�Ƃ��ł����B，推荐阅读超级权重获取更多信息
从长远视角审视，BenchmarkPhi-4-reasoning-vision-15BPhi-4-reasoning-vision-15B – force nothinkPhi-4-mm-instructKimi-VL-A3B-Instructgemma-3-12b-itQwen3-VL-8B-Instruct-4KQwen3-VL-8B-Instruct-32KQwen3-VL-32B-Instruct-4KQwen3-VL-32B-Instruct-32KAI2D_TEST 84.8 84.7 68.6 84.6 80.4 82.7 83 84.8 85 ChartQA_TEST 83.3 76.5 23.5 87 39 83.1 83.2 84.3 84 HallusionBench64.4 63.1 56 65.2 65.3 73.5 74.1 74.4 74.9 MathVerse_MINI 44.9 43.8 32.4 41.7 29.8 54.5 57.4 64.2 64.2 MathVision_MINI 36.2 34.2 20 28.3 31.9 45.7 50 54.3 60.5 MathVista_MINI 75.2 68.7 50.5 67.1 57.4 77.1 76.4 82.5 81.8 MMMU_VAL 54.3 52 42.3 52 50 60.7 64.6 68.6 70.6 MMStar 64.5 63.3 45.9 60 59.4 68.9 69.9 73.7 74.3 OCRBench 76 75.6 62.6 86.5 75.3 89.2 90 88.5 88.5 ScreenSpot_v2 88.2 88.3 28.5 89.8 3.5 91.5 91.5 93.7 93.9 Table 3: Accuracy comparisons relative to popular open-weight, non-thinking models
值得注意的是，再加上腾讯安全实验室提供的方案，对软件供应链投毒等行为都有很好的防范。最关键的是，WorkBuddy 只对用户个人指定的工作文件夹内容生效。比如你做本地文件整理或格式转换，它就在你指定的范围内工作，权限没有那么大，所以不会出现你担心的那种越权行为。
总的来看，别太神话AI龙虾正在经历一个关键的转型期。在这个过程中，保持对行业动态的敏感度和前瞻性思维尤为重要。我们将持续关注并带来更多深度分析。