Z.ai (Zhipu AI) · open · 2026-06
GLM-5.2
not ranked n/a
the takeThe strongest open-weights model on the independent index, MIT-licensed, and a genuine problem for the closed labs. Off the board here only because Z.ai skipped SWE-bench Verified, MMLU-Pro and LiveCodeBench, so I cannot place it fairly against the rest.
Not ranked: Z.ai (Zhipu AI) doesn't publish enough comparable public benchmarks to place it fairly. How ranking works.
GPQA Diamond
91.2
MMLU-Pro
—
AIME (math)
99.2
- Context
- 1M
- Input
- $1.4/M
- Output
- $4.4/M
- Speed
- —
- Modality
- text
- Top open-weights model on the AA index
- Leads open models on agentic coding
- Strong math/science, 1M context, MIT
- Text-only
- Skips the standard comparable benchmarks
- ~753B MoE needs serious hardware to self-host