Skip to content

Update infra: gpt-4.1 models, AVM 0.10.0, tier comments#70

Merged
diberry merged 16 commits into
Azure-Samples:mainfrom
diberry:infra/update-openai-models-tier
Apr 29, 2026
Merged

Update infra: gpt-4.1 models, AVM 0.10.0, tier comments#70
diberry merged 16 commits into
Azure-Samples:mainfrom
diberry:infra/update-openai-models-tier

Conversation

@diberry

@diberry diberry commented Apr 27, 2026

Copy link
Copy Markdown
Collaborator

Changes

  • Chat model: gpt-4o-mini → gpt-4.1-mini (version 2025-04-14)
  • Synth model: gpt-4o → gpt-4.1 (version 2025-04-14)
  • AVM version: cognitive-services/account 0.7.1 → 0.14.0
  • API version: 2024-08-01-preview → 2025-04-01-preview (chat + synth models)
  • Tier: Standard SKU (unchanged, documented with comments)
  • READMEs: Added azd env get-values > .env instructions to all sample READMEs
  • azure.yaml: Removed postprovision/postdown hooks (no auto .env creation)

What's NOT changed

  • Embedding model (text-embedding-3-small) — unchanged
  • All outputs — unchanged
  • Location param — unchanged

Why

The gpt-4o-mini Standard deployment was deprecated (2026-03-31). This updates to the gpt-4.1 model family and bumps the AVM module version. Comments help future contributors understand how to switch between Standard and GlobalStandard tiers.

Reference

Model availability: https://learn.microsoft.com/azure/ai-services/openai/concepts/models?tabs=python-secure%2Cglobal-standard%2Cstandard-chat-completions#models-by-deployment-type

diberry and others added 5 commits April 27, 2026 14:55
- Chat: gpt-4o-mini → gpt-4.1-mini (2025-04-14)
- Synth: gpt-4o → gpt-4.1 (2025-04-14)
- AVM: cognitive-services/account 0.7.1 → 0.10.0
- Add tier selection comments (Standard ↔ GlobalStandard)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
OpenAI can now be deployed to a different region than other resources.
Supported OpenAI regions: eastus, eastus2, eastus3, westus, westus2, westus3, northeurope, swedencentral.
Defaults to main location if not specified.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Reads AZURE_OPENAI_LOCATION env var, falls back to AZURE_LOCATION.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Promote chatModelName, chatModelType, synthModelName, synthModelType,
embeddingModelName, embeddingModelType from hardcoded vars to params
with env var overrides in bicepparam.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
GlobalStandard has 0 quota on BAMI_DIBERRY_2 subscription.
Standard tier has available quota for gpt-4.1 (1000K) and
gpt-4.1-mini (5000K) in eastus2.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
diberry and others added 4 commits April 28, 2026 10:37
Document azd env set commands for customizing OpenAI
location, model names, versions, and deployment SKUs
before running azd up.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
DocumentDB (MongoDB vCore) allows vector indexes to be created or
dropped on existing collections at any time. Added note to root README
and all sample READMEs.

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

- managed-identity AVM 0.4.0 → 0.5.0
- cognitive-services AVM 0.10.0 → 0.14.0
- API versions → 2025-04-01-preview (chat + synth)
- Remove hardcoded password, use readEnvironmentVariable

This makes PR Azure-Samples#70 a complete superset of PR Azure-Samples#59.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
diberry and others added 6 commits April 28, 2026 12:44
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- agent-go: Add 'Deploy Azure resources' section with azd up + azd env get-values > .env
- agent-typescript: Update TIP to use 'azd env get-values > .env' export command

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Document common failure modes (region availability, deployment type mismatch,
quota limits) and remediation steps for Azure OpenAI provisioning issues.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Empty default violated @minlength(8) constraint, breaking azd down/up.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
@diberry diberry closed this Apr 28, 2026
@diberry diberry reopened this Apr 28, 2026
@diberry diberry merged commit 9db626a into Azure-Samples:main Apr 29, 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.

1 participant