Z.ai Launches GLM-5.2 With a Usable 1M-Token Context, Two Thinking-Effort Levels, and No Benchmarks at Launch

- Z.ai dropped GLM-5.2 with a 1M-token context window, effectively letting coding agents hold entire repositories in memory without the usual summarization amnesia. It’s compatible with Claude Code and Cline, offering High and Max thinking effort for complex refactors. The catch? Zero benchmarks at launch. It’s like selling a hash manifest with no PoD seal—useful, but you’re flying blind on actual performance. Z.ai promises MIT weights next week, so the open-source crowd can finally verify if this is genuine infrastructure or just another hype cycle. Until then, it’s a powerful tool wrapped in marketing smoke. Don’t bet your stack-eye on unverified metrics; let the code speak when the weights drop.