Introducing GLM 5.3 on Amazon Bedrock
753B MoE for coding agents: GLM 5.3 hits Bedrock with prompt caching and OpenAI-compatible APIs.
Why it matters
A large mixture-of-experts model purpose-built for agentic tasks is now available in Bedrock. Practitioners can invoke it via standard APIs, reduce latency and cost with prompt caching, and test it with open-source agent frameworks. The headline is the product availability and integration path, not frontier capability.
The key facts
12 to knowGLM 5.3: 753B-parameter mixture-of-experts model
Optimized for coding and long-horizon agentic tasks
Available on Amazon Bedrock
OpenAI-compatible API support
Prompt caching for latency and cost reduction
Open-source Strix agent framework for authorized security testing
Published Oct 5, 2026
Built for coding and long-horizon agentic tasks
OpenAI-compatible APIs supported
Prompt caching for cost and latency reduction
Strix agent (open-source) included for authorized security testing
Published: October 5, 2026
The story so far
Earlier coverage of this storyline
- Introducing Anthropic models on Amazon Bedrock for in-region inference in Seoul and SingaporeAWS Machine Learning Blog
- This story
Go to the source
AWS Machine Learning Blogaws.amazon.com
Publisher excerpt: 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…