Tested AI Paper Paraphrasing Prompt: Weipu Detection Rate Drops to 6%
A custom prompt successfully reduced AI paper detection rates to 6% on Weipu and near 0% on Tencent Zhuque in practical tests.
A custom prompt successfully reduced AI paper detection rates to 6% on Weipu and near 0% on Tencent Zhuque in practical tests.
AI generates 3D Ace Combat game using MiniMax-M2.1 - practical test shows potential for creative programming and rapid prototyping.

ECS / OSS / CDN / 云数据库一站采购,常用云资源集中选配;新用户与续费均有专场优惠,适合个人开发者与小团队长期使用。
Testing reveals Google's Gemini 3 Pro High and Low versions may be identical, hitting quota limits at the same time.
Minimax opens M2.1 beta access for programming tests. AI enthusiasts invited to test and compare model performance.
Gemini 3 Pro outperformed other AIs in a matchmaking test, showing superior logic, especially on counter-intuitive social problems.
GPT-5.2-Codex test shows 50% speed increase over GPT-5.2, with similar output quality but no capability improvements.
ChatGPT struggles with C language trap questions while Gemini and DeepSeek excel, revealing AI performance differences in programming tasks.
Google's Gemini 3 Flash model is now available via CLI2API. Tests show smooth integration with CherryStudio, though thinking mode needs manual enabling.
Google's Gemini 3 Flash launches with 50% faster response times and lower costs than its Pro predecessor.
Xiaomi MiMo-V2-Flash excels in AI programming test, outperforming competitors in LeetCode challenge.