First Time Encountering Zhipu Model's Review Gate

First, the task was merely to translate Chinese to English for a nodejs app that converts a mermaid into an image. This is a pure tool, and the task was only translation. So far, I have encountered the following issues while using the Zhipu model: 1. Extremely low concurrency limits, the more you need to use it, the lower the concurrency 2. Low token generation speed 3….
Tags:
Categories:

First, the task was merely to translate Chinese to English for a nodejs app that converts a mermaid into an image. This is a pure tool, and the task was only translation. So far, I have encountered the following issues while using the Zhipu model:

  1. Extremely low concurrency limits, the more you need to use it, the lower the concurrency
  2. Low token generation speed
  3. Multi-turn tool calls falling into an infinite loop
  4. Risk control, with no documentation explaining its impact
  5. Sensitive content gate, with no definition of what content is sensitive In the glm4.5 version, I recommended Zhipu’s plans, never expecting that the user experience has been continuously declining since then. Always being marketed to and then spending money to test, wasting a lot of time. Given Zhipu’s large-scale overselling and its desperate rate-limiting measures, the poor experience issue is likely to persist for a while. I plan to check how Zhipu develops in a year, and for one year I will not believe any recommendations about Zhipu.

First Time Encountering Zhipu Model’s Review Gate Figure 1

First Time Encountering Zhipu Model’s Review Gate Figure 2