Introducing GLM 5.3 on Amazon Bedrock
GLM 5.3 from Z.ai is now available on Amazon Bedrock: a 753B-parameter mixture-of-experts model built for coding and long-horizon agentic tasks. Learn how to invoke it with the OpenAI-compatible APIs, cut cost and latency with prompt caching, and run an authorized security test with the open-source Strix agent.
The Z.ai-developed GLM 5.3 model, a massive 753 billion parameter mixture-of-experts model, has been made available on Amazon Bedrock for coding and long-horizon agentic tasks. This new model boasts significant improvements over its predecessor GLM 5, particularly in coding benchmarks and security tasks. Amazon Bedrock offers fully managed APIs, cross-Region inference, prompt caching, and service tiers, eliminating the need for users to manage any infrastructure.
To utilize GLM 5.3, users can invoke it through OpenAI-compatible APIs or Amazon Bedrock's Invoke and Converse APIs, with prompt caching available for agentic workloads. The model's cross-Region inference profiles allow users to choose US or Global inference regions for processing. Service tiers such as Flex, Priority, and Standard provide flexibility in cost and speed optimization.
To get started, users can access GLM 5.3 on the Amazon Bedrock console without writing code or installing developer tools. Simply navigate to Amazon Bedrock, select Test Playground, choose GLM 5.3, and send prompts through the chat UI. For programmatic access, the AWS Command Line Interface (AWS CLI) credentials can be used to generate short-lived tokens, which are then used to call the Responses API from Python using the OpenAI Python SDK and the aws-bedrock-token-generator library.
This article demonstrates how to refactor a recursive Python function into an iterative one, showcasing the improved coding capabilities of GLM 5.3.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
