As stated in the title
ModelOC
gpt-5.5 OpenAI · 6066.pw
评级 A 评分 100 风险 低
通过 11警告 1不确定 4
测试时间: 2026-07-30 01:34 UTC
I tested this myself using GPT-OSS-120b, and it feels like there’s no intelligence at all.
There’s a contradiction with this platform: rich people can afford the computing power, so the platform is just a toy for them—yet if they’re already rich, wouldn’t they prefer buying official access? On the other hand, folks without money or resources simply can’t play with this ‘toy’ at all. Maybe the developers never intended to include them in the first place, so at least it’s not taking advantage of poor people—which is a good thing.
I’m talking about detecting authenticity, not selling it. I just suddenly wanted to test today whether this model could pass, and it turned out to get a high score.
This really is a bit much; I have no idea how they even detect water injection.
可能评判内容没更新？
第三方AI检测图一乐，只能看你的使用经验判断
For reference only
应该蛮准的把
