fix: remove unstable stability flag for Azure models by steebchen · Pull Request #1920 · theopenco/llmgateway

steebchen · 2026-03-29T07:34:48Z

Summary by CodeRabbit

Chores
- Updated Azure provider configuration for GPT-5, GPT-5 Mini, and GPT-5 Nano models with pricing adjustments.

coderabbitai · 2026-03-29T07:35:06Z

Walkthrough

Updates Azure provider entries for GPT-5 model variants (gpt-5, gpt-5-mini, gpt-5-nano) by removing the stability: "unstable" property and adding discount: 0.3. No other pricing fields, context sizes, or capabilities were modified.

Changes

Cohort / File(s)	Summary
Azure GPT-5 Provider Metadata `packages/models/src/models/openai.ts`	Removed `stability: "unstable"` and added `discount: 0.3` to Azure provider entries for three GPT-5 model variants (gpt-5, gpt-5-mini, gpt-5-nano).

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Possibly related PRs

PR #1367: Adds Azure provider mappings and marks GPT-5 variants as unstable, establishing the baseline entries modified in this PR.
PR #1183: Directly adjusts the same Azure GPT-5 entries, adding stability flags that this PR then removes with discount configuration.
PR #1538: Removes stability: "unstable" from specific model provider entries, following the same metadata cleanup pattern.

Suggested reviewers

smakosh
kmk142789

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main change: removing the unstable stability flag from Azure model entries (gpt-5, gpt-5-mini, gpt-5-nano) and replacing it with a discount field.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

Create stacked PR
Commit on current branch

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch feat/azure

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

coderabbitai

🧹 Nitpick comments (1)

packages/models/src/models/openai.ts (1)
839-866: Verify if other Azure models should also have stability flag removed.

The changes to gpt-5, gpt-5-mini, and gpt-5-nano look correct and consistent. However, several other Azure model providers still retain stability: "unstable", including:

Codex variants: gpt-5.1-codex, gpt-5.1-codex-mini, gpt-5.2-codex, gpt-5.3-codex

Chat variants: gpt-5.2-chat-latest, gpt-5.3-chat-latest

gpt-5.2-pro

If these models should also be marked as stable on Azure, they could be updated in a follow-up. If this scoping is intentional (e.g., only base variants are stable), no action needed.

Also applies to: 905-932
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.

In `@packages/models/src/models/openai.ts` around lines 839 - 866, The PR removed
the stability flag for azure base gpt-5 variants but left several Azure model
entries still marked stability: "unstable"; review the Azure model objects for
gpt-5.1-codex, gpt-5.1-codex-mini, gpt-5.2-codex, gpt-5.3-codex,
gpt-5.2-chat-latest, gpt-5.3-chat-latest, and gpt-5.2-pro in
packages/models/src/models/openai.ts and either remove the stability: "unstable"
property or change it to the intended value (e.g., delete the key to make them
stable) so the Azure entries are consistent with the gpt-5, gpt-5-mini, and
gpt-5-nano updates.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@packages/models/src/models/openai.ts`:
- Around line 839-866: The PR removed the stability flag for azure base gpt-5
variants but left several Azure model entries still marked stability:
"unstable"; review the Azure model objects for gpt-5.1-codex,
gpt-5.1-codex-mini, gpt-5.2-codex, gpt-5.3-codex, gpt-5.2-chat-latest,
gpt-5.3-chat-latest, and gpt-5.2-pro in packages/models/src/models/openai.ts and
either remove the stability: "unstable" property or change it to the intended
value (e.g., delete the key to make them stable) so the Azure entries are
consistent with the gpt-5, gpt-5-mini, and gpt-5-nano updates.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 7321898c-22f0-49db-98cd-1a9f21c4ea03

📥 Commits

Reviewing files that changed from the base of the PR and between 4e6335d and cbdefcc.

📒 Files selected for processing (1)

packages/models/src/models/openai.ts

Copilot

Pull request overview

Removes the stability: "unstable" flag from select Azure OpenAI model mappings so they’re no longer treated as unstable by downstream model-selection/filtering logic.

Changes:

Removed provider-level stability: "unstable" from Azure mappings for gpt-5, gpt-5-mini, and gpt-5-nano.
Added discount: 0.3 to those same Azure mappings (affecting effective price calculations).

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

Copilot · 2026-03-29T07:37:55Z

packages/models/src/models/openai.ts

 				test: "skip",
 				providerId: "azure",
-				stability: "unstable" as const,
 				modelName: "gpt-5",
 				inputPrice: 1.25 / 1e6,
 				outputPrice: 10.0 / 1e6,
+				discount: 0.3,
 				cachedInputPrice: 0.125 / 1e6,


PR title indicates removing the unstable stability flag for Azure mappings, but this change only removes it for the GPT-5 / GPT-5-mini / GPT-5-nano Azure entries. There are still multiple Azure mappings in this file that set stability: "unstable" (e.g. around openai.ts:1138, 1249, 1297, 1566, 1615, 1664). If the intent is to stop treating Azure as unstable, please remove/update those remaining flags as well (or adjust the PR title/scope to match).

Copilot · 2026-03-29T07:37:56Z

packages/models/src/models/openai.ts

 				test: "skip",
 				providerId: "azure",
-				stability: "unstable" as const,
 				modelName: "gpt-5-mini",
 				inputPrice: 0.25 / 1e6,
 				outputPrice: 2 / 1e6,
 				cachedInputPrice: 0.025 / 1e6,
+				discount: 0.3,
 				requestPrice: 0,


This hunk introduces discount: 0.3 for the Azure mappings while also removing stability. That changes effective pricing used by provider/model selection (discount affects cheapest-provider logic), which is a behavior change beyond "remove unstable stability flag". If Azure pricing should remain identical to OpenAI, drop the discount; otherwise please document/justify the Azure discount (and consider applying it consistently across all Azure mappings for this family).

Copilot · 2026-03-29T07:37:56Z

packages/models/src/models/openai.ts

+				discount: 0.3,
 				cachedInputPrice: 0.125 / 1e6,


discount is placed inconsistently relative to the other pricing fields in the Azure GPT-5 mappings (e.g. here it appears before cachedInputPrice, while elsewhere in this file discounts typically come after cachedInputPrice). Please reorder fields consistently (keep pricing-related fields grouped in the same order) to reduce maintenance mistakes when updating prices later.

Suggested change

discount: 0.3,

cachedInputPrice: 0.125 / 1e6,

cachedInputPrice: 0.125 / 1e6,

discount: 0.3,

Copilot · 2026-03-29T07:37:56Z

packages/models/src/models/openai.ts

+				discount: 0.3,
 				inputPrice: 0.05 / 1e6,
 				outputPrice: 0.4 / 1e6,
 				cachedInputPrice: 0.005 / 1e6,


In this Azure mapping, discount is inserted before inputPrice (unlike other entries where discount is grouped with the pricing fields, typically after cachedInputPrice). Please keep a consistent field ordering for pricing attributes to avoid overlooking discount when updating token prices.

Suggested change

discount: 0.3,

inputPrice: 0.05 / 1e6,

outputPrice: 0.4 / 1e6,

cachedInputPrice: 0.005 / 1e6,

inputPrice: 0.05 / 1e6,

outputPrice: 0.4 / 1e6,

cachedInputPrice: 0.005 / 1e6,

discount: 0.3,

fix: remove unstable stability flag for Azure models

cbdefcc

Copilot AI review requested due to automatic review settings March 29, 2026 07:34

Copilot started reviewing on behalf of steebchen March 29, 2026 07:35 View session

steebchen self-assigned this Mar 29, 2026

coderabbitai bot reviewed Mar 29, 2026

View reviewed changes

Copilot AI reviewed Mar 29, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix: remove unstable stability flag for Azure models#1920

fix: remove unstable stability flag for Azure models#1920
steebchen wants to merge 1 commit intomainfrom
feat/azure

steebchen commented Mar 29, 2026 •

edited by coderabbitai bot

Loading

Uh oh!

coderabbitai bot commented Mar 29, 2026 •

edited

Loading

Uh oh!

coderabbitai bot left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

steebchen commented Mar 29, 2026 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Uh oh!

coderabbitai bot commented Mar 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Possibly related PRs

Suggested reviewers

Uh oh!

coderabbitai bot left a comment

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

steebchen commented Mar 29, 2026 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Mar 29, 2026 •

edited

Loading