GLM-5.2 is currently the best open-weight model for development. It's not as good as the current American SOTA models, but if you wrote code with US SOTA models four months ago, you can write code with GLM-5.2 today.
DeepSeek 4 is a good model for many tasks, but I think it currently lacks the post-training required to become a genuinely great coding model.
DeepSeek 4 is a good model for many tasks, but I think it currently lacks the post-training required to become a genuinely great coding model.