GLM-5.3 API Launch: $1.40/$4.40 per 1M Tokens (2-min read)

The latest open-source frontier model GLM-5.3 from Chinese startup z.ai has been released via its official API. The model has demonstrated exceptional cybersecurity capabilities, including discovering previously unknown vulnerabilities in Cursor.

API pricing remains identical to its predecessor, GLM-5.2, at $1.40 per 1M input tokens and $4.40 per 1M output tokens. Cached input costs $0.26 per 1M tokens, and cache storage fees are temporarily free.

This allows developers to switch to GLM-5.3, which offers higher coding and long-horizon agent performance, while maintaining the same token unit prices as before. Currently, GLM Coding Plan subscribers can only use the OpenAI Chat Completions compatible protocol, and the release schedule and license for the model weights have not yet been finalized.