Skip to main content
Rase AI Intelligence Newsletter

AI infrastructure Model or AI service

Fixstars adds Japan-hosted GLM-5.3 inference for direct business customers

Fixstars added an optimized GLM-5.3 model to its Japan-hosted inference service, initially for businesses with direct contracts. [1][2]

Model access
GLM-5.3 offered to direct-contract business customers; gateway rollout planned [1][2]

What the model does

The provider quantizes Z.ai's existing open-weight model to NVIDIA's NVFP4 format and serves it through its managed GPU infrastructure. The November 20 release says GLM-5.3 access through the FastMetal gateway is planned for later; access to other existing models through that gateway does not establish GLM-5.3 availability there. [1][2]

Why it matters

Organizations evaluating where AI inputs are processed gain another hosting option for an existing model. This is a deployment and access update, not the invention or original release of GLM-5.3. [1][2]

Who can use it

Businesses seeking Japan-based inference and developers integrating tools that support OpenAI-compatible APIs. [1][2]

Key dates

  1. Fixstars adds GLM-5.3 to its Japan-hosted inference service [1][2]

Ways to use it

Direct-contract applicants receive endpoint details after application and review. Confirm the contracted model and routing rather than assuming a generic gateway endpoint stays in Japan. [1][2]

Model limitations

Fixstars' data-residency and no-persistent-input/output-retention statements are provider claims, not an independent audit. Reported speed and cost comparisons depend on the stated workload and date; this profile makes no universal fastest or cheapest claim. [1][2]

Sources and verification

Catalog evidence checked How we verify records

  1. Primary source Fixstars launches GLM-5.3 on Fixstars Inference Cloud
  2. Primary source Fixstars Inference Cloud: Fast, Japan-hosted inference for open-weight models

Related profiles

AI infrastructure

Commerce OIG review of advanced-compute diversion controls

U.S. Department of Commerce Office of Inspector General

The signed initiation memo sets out an examination of whether BIS designed and implemented controls to detect, prevent and respond to chip/server diversion through third countries and unauthorised access to overseas computing capacity.

High confidenceEvidence checked

AI infrastructure

Amazon SageMaker AI inference endpoint monitoring

Amazon Web Services

Notifications link to configuration recommendations in SageMaker Studio. Customers can test alternatives against recent traffic and compare results before applying a change; production endpoints remain untouched during these tests.

High confidenceEvidence checked

AI infrastructure

NVIDIA–Microsoft AI infrastructure collaboration

NVIDIA and Microsoft

The announcement covers Vera Rubin systems in Azure, NVIDIA models in Foundry, and additional edge and research integrations. It explicitly mixes currently running infrastructure, previews and capabilities still being developed.

High confidenceEvidence checked