chore: adds docs for azure model router

[sammaji]

Summary

Adds a dedicated documentation page for the Azure Model Router provider, explaining how Bifrost automatically falls back to Chat Completions when a model-router deployment is targeted via the Responses API.

Changes

• Added docs/providers/supported-providers/azure-model-router.mdx documenting the Azure model-router routing behavior, supported operations, a Mermaid flowchart illustrating the fallback logic, usage examples (REST and Go SDK), and known limitations.
• Registered the new page in docs/docs.json so it appears in the navigation between the Azure and Bedrock entries.

Type of change

• Bug fix
• Feature
• Refactor
• Documentation
• Chore/CI

Affected areas

• Core (Go)
• Transports (HTTP)
• Providers/Integrations
• Plugins
• UI (React)
• Docs

How to test

Navigate to the Azure Model Router page in the rendered docs and verify:

• The page appears in the sidebar between Azure and Bedrock.
• The Mermaid flowchart renders correctly.
• All code examples and notes display as expected.

Breaking changes

• Yes
• No

Related issues

Security considerations

None. This is a documentation-only change with no impact on auth, secrets, or runtime behavior.

Checklist

• I read docs/contributing/README.md and followed the guidelines
• I added/updated tests where appropriate
• I updated documentation where needed
• I verified builds succeed (Go and UI)
• I verified the CI pipeline passes locally if applicable

[coderabbitai]

📝 Walkthrough

Summary by CodeRabbit

• Documentation
  • Added documentation for using Azure OpenAI model-router deployments.
  • Explained supported Chat Completions usage and the configuration workaround for Responses.
  • Included Web UI, API, and configuration file setup examples.
  • Added endpoint usage guidance, limitations, and related documentation links.
  • Updated the provider documentation navigation to include the new page.

Walkthrough

Adds documentation for Azure OpenAI model-router deployments, including Chat Completions support, Responses configuration through Azure AI Foundry, usage examples, limitations, related documentation, and navigation.

Changes

Azure Model Router documentation

Layer / File(s)	Summary
Document model-router setup and usage docs/providers/supported-providers/azure-model-router.mdx	Adds the model-router overview, supported-operations table, Azure AI Foundry configuration steps, Responses request example, limitations, and related documentation links.
Add supported-provider navigation entry docs/docs.json	Adds the Azure model-router page to the supported-provider navigation after Azure.

Estimated code review effort: 2 (Simple) | ~10 minutes

Suggested reviewers: bearts, pratham-mishra04

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and accurately describes the documentation addition for the Azure model router.
Description check	✅ Passed	The description covers the required sections and is mostly complete, with only minor non-critical items left blank.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch 07-08-chore_adds_docs_for_azure_model_router

Warning

There were issues while running some tools. Please review the errors and either fix the tool's configuration or disable the tool if it's a critical failure.

🔧 Checkov (3.3.8)

docs/docs.json

Traceback (most recent call last):
File "/usr/local/bin/checkov", line 2, in
from checkov.main import Checkov
ModuleNotFoundError: No module named 'checkov'

---

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

[sammaji]

• fix: adds created timestamp for openai text completions #5407
• chore: adds docs for azure model router #5175 👈 (View in Graphite)
• dev

This stack of pull requests is managed by Graphite. Learn more about stacking.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[greptile-apps]

Confidence Score: 4/5

The raw-response logging guidance needs correction before merging.

• The navigation entry and main configuration example match existing documentation patterns.
• The logging note can cause users to persist raw provider data when they intend not to store it.

docs/providers/supported-providers/azure-model-router.mdx

Security Review

The raw-response note can cause sensitive provider data to be persisted because it recommends setting the storage option when the stated goal is to avoid storage.

Important Files Changed

Filename	Overview
docs/docs.json	Adds the new page to the Supported Providers navigation with a matching route.
docs/providers/supported-providers/azure-model-router.mdx	Adds the provider guide, but reverses the meaning of the raw-data storage option.

_{Reviews (1): Last reviewed commit: "chore: adds docs for azure model router" | Re-trigger Greptile}

[greptile-apps]

Raw Logging Advice Is Reversed

When a user wants to inspect raw responses without persisting them, this note tells them to set Store Raw Request/Response, which enables internal log persistence rather than preventing it. This can store sensitive provider data against the user's intent; the option should remain disabled, and the corrected link should use the generated section fragment.

Suggested change

	If you need to inspect provider-specific extra parameters, enable [Send Back Raw Response](/providers/request-options#send-back-raw-response). If you do not want those raw bytes persisted in logs, also set [Store Raw Request/Response](/providers/request-options#store-raw-request/response).
	If you need to inspect provider-specific extra parameters, enable [Send Back Raw Response](/providers/request-options#send-back-raw-response). If you do not want those raw bytes persisted in logs, leave [Store Raw Request/Response](/providers/request-options#store-raw-requestresponse) disabled.

[coderabbitai]

Actionable comments posted: 3

🧹 Nitpick comments (1)

docs/providers/supported-providers/azure-model-router.mdx (1)

113-115: 🔒 Security & Privacy | 🔵 Trivial | ⚡ Quick win

Clarify the independence of the raw-response controls.

Use a <Warning> block to distinguish returning provider-specific data from persisting raw requests/responses, and state which setting must be disabled to prevent persistence. The current <Note> can imply that enabling one control governs the other.

As per path instructions, config-flag independence should be called out explicitly with a Warning.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/providers/supported-providers/azure-model-router.mdx` around lines 113 -
115, The raw-response guidance in the Note should explicitly distinguish
returning provider-specific data from persisting raw requests/responses. Replace
the Note with a Warning that explains these controls are independent and states
that Store Raw Request/Response must be disabled to prevent persistence, while
retaining the Send Back Raw Response setting for inspecting provider-specific
parameters.

Source: Path instructions

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@docs/providers/supported-providers/azure-model-router.mdx`:
- Around line 100-111: Add a Go SDK usage example alongside the existing curl
example in the Usage section, following the repository’s established SDK
conventions and using the same azure-foundry provider, model-router deployment,
Responses endpoint, and input prompt.
- Around line 84-89: Update the request_path_overrides entry in the field table
to mark it as required for the documented Azure Foundry Responses setup, and
clarify that both responses and responses_stream overrides must be provided to
route requests to the Responses endpoint. Leave the other field requirements
unchanged.
- Line 30: Update the Azure AI Foundry documentation to use the
*.services.ai.azure.com/api/projects/<project> hostname throughout the prose, UI
example, and JSON request_path_overrides. In the custom provider example, remove
base_url or explicitly state that the full-URL responses and responses_stream
overrides bypass it.

---

Nitpick comments:
In `@docs/providers/supported-providers/azure-model-router.mdx`:
- Around line 113-115: The raw-response guidance in the Note should explicitly
distinguish returning provider-specific data from persisting raw
requests/responses. Replace the Note with a Warning that explains these controls
are independent and states that Store Raw Request/Response must be disabled to
prevent persistence, while retaining the Send Back Raw Response setting for
inspecting provider-specific parameters.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: b4f3473e-7b67-4b57-9543-85dc64c2992e

📥 Commits

Reviewing files that changed from the base of the PR and between 878938e and 0fa6e8a.

📒 Files selected for processing (2)

• docs/docs.json
• docs/providers/supported-providers/azure-model-router.mdx

[coderabbitai]

🎯 Functional Correctness | 🟠 Major | ⚡ Quick win

🧩 Analysis chain

🏁 Script executed:

#!/bin/sh
set -eu
curl -fsSL 'https://learn.microsoft.com/en-us/rest/api/microsoft-foundry/aiproject' |
  grep -F 'services.ai.azure.com/api/projects'

Repository: maximhq/bifrost

Length of output: 35769

---

🏁 Script executed:

#!/bin/sh
set -eu

file='docs/providers/supported-providers/azure-model-router.mdx'

# Show the relevant sections with line numbers.
sed -n '1,180p' "$file" | cat -n

echo
echo '---'
echo 'Occurrences of ai.azure.com / services.ai.azure.com in this file:'
grep -nE 'ai\.azure\.com|services\.ai\.azure\.com|base_url|request_path_overrides|responses_stream|responses' "$file" || true

Repository: maximhq/bifrost

Length of output: 9228

---

Use the Foundry project hostname

The examples still point at *.ai.azure.com/api/projects/..., but the Foundry project endpoint uses *.services.ai.azure.com/api/projects/<project>. Update the prose, UI example, and JSON overrides, and either remove the base_url example or call out that the full-URL overrides ignore it.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/providers/supported-providers/azure-model-router.mdx` at line 30, Update
the Azure AI Foundry documentation to use the
*.services.ai.azure.com/api/projects/<project> hostname throughout the prose, UI
example, and JSON request_path_overrides. In the custom provider example, remove
base_url or explicitly state that the full-URL responses and responses_stream
overrides bypass it.

[coderabbitai]

🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win

Mark the Responses override as required for this setup.

request_path_overrides is documented as optional, but the page states that the full override is what routes requests to the Azure Foundry Responses endpoint. Without the responses and responses_stream overrides, this workaround has no documented path to that endpoint.

As per path instructions, the field table should make required fields explicit for the documented scenario.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/providers/supported-providers/azure-model-router.mdx` around lines 84 -
89, Update the request_path_overrides entry in the field table to mark it as
required for the documented Azure Foundry Responses setup, and clarify that both
responses and responses_stream overrides must be provided to route requests to
the Responses endpoint. Leave the other field requirements unchanged.

Source: Path instructions

[coderabbitai]

📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win

Add the promised Go SDK example.

The PR objectives call for REST and Go SDK usage examples, but this section only documents curl. Add a Go SDK example using the repository’s established SDK conventions.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/providers/supported-providers/azure-model-router.mdx` around lines 100 -
111, Add a Go SDK usage example alongside the existing curl example in the Usage
section, following the repository’s established SDK conventions and using the
same azure-foundry provider, model-router deployment, Responses endpoint, and
input prompt.

[akshaydeo]

Merge activity

• Jul 21, 2:59 PM UTC: A user started a stack merge that includes this pull request via Graphite.
• Jul 21, 2:59 PM UTC: @akshaydeo merged this pull request with Graphite.