Z.ai released GLM-5.2, a 753-billion parameter open-weights model designed for autonomous coding tasks. The model outperforms OpenAI's GPT-5.5 on multiple long-horizon coding benchmarks while costing one-sixth as much to run.
GLM-5.2 launched today across Hugging Face, Z.ai's API, and over 20 third-party coding environments. The model supports a stable 1-million-token context window, enabling it to handle extended code generation and debugging workflows. Enterprise subscriptions start at $12.60 monthly.
The performance gains matter for practical engineering work. Long-horizon coding tasks require models to maintain context across thousands of lines of code, understand complex dependencies, and generate multi-step solutions without losing track of previous decisions. GLM-5.2's architecture addresses this directly with its extended context window and specific training for coding scenarios.
Cost efficiency shifts economics for development teams. Running GLM-5.2 costs roughly one-sixth of GPT-5.5 per inference, making extended context windows financially viable for continuous integration pipelines, automated code review, and real-time pair programming assistance. Smaller teams and startups can now afford sophisticated AI coding tools previously limited to well-funded enterprises.
Z.ai's decision to open-source the weights under an unrestricted license removes vendor lock-in concerns. Teams can self-host GLM-5.2 on their own infrastructure, audit the model for security risks, and fine-tune it for specialized codebases without relying on external APIs.
The benchmark results signal a shift in AI development priorities. Rather than chasing general-purpose intelligence on academic benchmarks, focused models designed for specific domains are delivering measurable advantages in real work. GLM-5.2 succeeds because it trades breadth for depth in