OpenRouter lands fast GLM-5.2 endpoints, nitro routing

OpenRouter's new dynamic routing options and high-speed endpoints make running the flagship open-weights GLM-5.2 model significantly faster and more reliable.

• –Dynamic routing via the ':nitro' suffix solves the provider availability and speed volatility problem for production AI agents.
• –The addition of Wafer and Fireworks AI fast variants introduces healthy competition among inference providers, driving down latency and costs.
• –GLM-5.2's 1M-token context window makes high-speed endpoints crucial for developers running long-horizon coding tasks and multi-step workflows.
• –Using the unified ':nitro' endpoint prevents vendor lock-in and eliminates the need for manual fallback logic in developer codebases.