Skip to content

fix: gptmodel update - #678

Merged
Roopan-Microsoft merged 3 commits into
mainfrom
psl-gptmodel-update
Jul 20, 2026
Merged

fix: gptmodel update#678
Roopan-Microsoft merged 3 commits into
mainfrom
psl-gptmodel-update

Conversation

@Shubhangi-Microsoft

@Shubhangi-Microsoft Shubhangi-Microsoft commented Jul 16, 2026

Copy link
Copy Markdown
Contributor

This pull request updates the deployment to use the newer gpt-5-mini model instead of gpt-4.1-mini throughout the codebase, infrastructure, and documentation. It also updates the default model version, capacity recommendations, and related parameters to align with the new model. Additionally, minor improvements were made to resource deployment scripts and code owners.

Model Upgrade and Configuration Changes:

  • Updated all references, parameters, and defaults from gpt-4.1-mini to gpt-5-mini in the deployment scripts (infra/main.bicep, infra/main.json, Deployment/checkquota.ps1, Deployment/quota_check_params.sh, .github/workflows/CI.yml) and documentation (docs/AVMPostDeploymentGuide.md, docs/CustomizingAzdParameters.md, docs/QuotaCheck.md). [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
  • Updated the default GPT model version from 2025-04-14 to 2025-08-07 in all relevant configuration files and documentation. [1] [2] [3] [4]

Documentation Updates:

  • Revised documentation to reflect the new model name, version, and recommended capacity for quota checks and deployment guidance. [1] [2] [3] [4] [5]

Infrastructure and Deployment Improvements:

  • Added availabilityZones: [] to the managed cluster configuration to avoid issues in regions/subscriptions without availability zones. [1] [2]
  • Fixed the dependency order for DNS zones in the generated ARM template to ensure proper deployment sequencing.

Maintenance:

  • Corrected usernames in the .github/CODEOWNERS file for consistency.This pull request makes several important updates to transition the default GPT model from gpt-4.1-mini to gpt-5-mini across the codebase, documentation, infrastructure templates, and deployment scripts. In addition, it introduces a new AKS (Azure Kubernetes Service) module and updates the default VM SKU for agent pools. These changes ensure that deployments and documentation are aligned with the latest model versions and infrastructure best practices.

Key changes:

Model Upgrade and Configuration

  • Updated the default GPT model from gpt-4.1-mini to gpt-5-mini throughout infrastructure templates (infra/main.bicep, infra/main.json), deployment scripts (Deployment/checkquota.ps1, Deployment/quota_check_params.sh), CI workflows (.github/workflows/CI.yml), and documentation. The default model version was also updated from 2025-04-14 to 2025-08-07. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

Infrastructure Enhancements

  • Introduced a new AKS (Azure Kubernetes Service) wrapper module in infra/avm/modules/compute/kubernetes.bicep to standardize and simplify AKS cluster deployments with configurable parameters and outputs, including identity principal IDs.
  • Updated the default VM SKU for the AKS agent pool from Standard_D4ds_v5 to Standard_D4ds_v6 and explicitly disabled zonal placement to improve compatibility across regions. [1] [2]

Documentation Updates

  • Revised documentation to reflect the new default GPT model, version, and recommended quota/capacity guidelines. This includes parameter tables, quota check instructions, and deployment guides. [1] [2] [3] [4] [5]

Maintenance and Minor Fixes

  • Corrected user handles in the .github/CODEOWNERS file for consistency.
  • Minor reordering in deployment dependencies for improved reliability in infra/main.json.
  • Updated Bicep template hash to reflect infrastructure changes.## Purpose
  • ...

Does this introduce a breaking change?

  • Yes
  • No

Golden Path Validation

  • I have tested the primary workflows (the "golden path") to ensure they function correctly without errors.

Deployment Validation

  • I have validated the deployment process successfully and all services are running as expected with this change.

What to Check

Verify that the following are valid

  • ...

Other Information

Shubhangi-Microsoft and others added 2 commits July 16, 2026 10:43
- Replace deprecated gpt-4.1-mini (2025-04-14) with gpt-5-mini (2025-08-07) in main.bicep/main.json, CI.yml, quota scripts and docs

- Set AKS agent pool availabilityZones=[] to avoid AvailabilityZoneNotSupported in regions/subs without zones

- Change AKS VM size Standard_D4ds_v5 -> Standard_D4ds_v6 (v5 not allowed in subscription)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR updates the solution’s default Azure OpenAI model configuration from gpt-4.1-mini to gpt-5-mini (including model version), and makes related infrastructure and documentation adjustments. It also updates AKS infrastructure defaults (agent pool VM SKU and zonal placement behavior) and adds a new AKS wrapper module under infra/avm/modules.

Changes:

  • Switch default GPT model to gpt-5-mini and version to 2025-08-07 across infra templates, CI, scripts, and docs.
  • Update AKS agent pool default VM SKU to Standard_D4ds_v6 and explicitly disable availability zones in the main infra template.
  • Add a new AKS AVM wrapper module (infra/avm/modules/compute/kubernetes.bicep) with parameters and outputs.

Reviewed changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 4 comments.

Show a summary per file
File Description
infra/main.json Updates default GPT model/version, quota usageName, AKS agent pool VM SKU, and agent pool zonal placement configuration.
infra/main.bicep Updates default GPT model/version and AKS agent pool defaults (SKU + availabilityZones: []).
infra/avm/modules/compute/kubernetes.bicep Adds a new AKS wrapper module around the AVM managed-cluster module with parameters/outputs.
docs/QuotaCheck.md Updates quota-check documentation examples to the new default GPT model.
docs/CustomizingAzdParameters.md Updates documented default environment parameters for GPT model name/version.
docs/AVMPostDeploymentGuide.md Updates recommended quota table to the new default GPT model.
Deployment/quota_check_params.sh Updates default model-capacity pair to gpt-5-mini.
Deployment/checkquota.ps1 Updates quota lookup key to OpenAI.GlobalStandard.gpt-5-mini.
.github/workflows/CI.yml Updates CI deployment parameters for GPT model name/version.
.github/CODEOWNERS Updates CODEOWNERS handles.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread infra/avm/modules/compute/kubernetes.bicep Outdated
Comment thread infra/avm/modules/compute/kubernetes.bicep Outdated
Comment thread docs/QuotaCheck.md
Comment thread infra/avm/modules/compute/kubernetes.bicep Outdated
Copilot AI review requested due to automatic review settings July 17, 2026 12:04

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Copilot reviewed 9 out of 9 changed files in this pull request and generated 1 comment.

Comment thread infra/main.bicep
@Roopan-Microsoft
Roopan-Microsoft merged commit d0f9494 into main Jul 20, 2026
9 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants