via
Z.aiKey points
- GLM-5.2 has claimed the top spot on the Artificial Analysis Intelligence Index with a score of 51.
- The model features a massive 1 million token context window and is released under the MIT license.
- Despite having the same parameter count as its predecessor, its intelligence index jumped by 11 points.
It's impressive to see how much efficiency can be squeezed out of the same architecture. GLM-5.2 maintains the 744B total / 40B active parameter split of the 5.1 version, yet the performance leap is significant, especially in agentic tasks where it's nearing high-reasoning frontier models.
However, the "intelligence" comes at a cost of verbosity. The data shows it's outputting an average of 43k tokens per task, with the vast majority being internal reasoning. While this explains the higher scores, the token efficiency is definitely a weak point. For developers, the MIT license is the real win here, making it highly accessible via multiple third-party providers. It's a powerful addition to the open-weight ecosystem, provided you can afford the output overhead.
Next Doubao Pro is Officially Here →
All Replies （0） #
No replies yet — be the first!