GLM-5.2 is the first open-weights model to exceed 80% on Terminal-Bench, outperforming all other open models and even Gemini

Core summary

One sentence to understand this update

GLM-5.2 achieved over 80% on Terminal-Bench, making it the first open-weights model to reach this milestone and surpass other open models and Gemini at a fraction of the cost.

Impact & opportunity

What this could mean

Builders can leverage GLM-5.2 for frontier-level performance in their applications with open-weights models, significantly reducing operational costs.