feat(azure): add URL auto-parser and improve settings UX

- Fix baseURL empty string fallback (pass undefined instead of "" to let SDK use env vars)
- Add parseAzureUrl() utility that extracts endpoint, deployment name, and API version from a full Azure deployment URL
- Integrate auto-parser into Azure settings: pasting a full URL auto-fills all fields
- Rename "Base URL" to "Azure Endpoint" to match Azure portal terminology
- Improve field descriptions for deployment name and API version
- Add 13 tests for URL parser (all passing)

Author: hannesrudolph

src/api/providers/azure.ts

@@ -1,6 +1,6 @@
 import { Anthropic } from "@anthropic-ai/sdk"
 import { createAzure } from "@ai-sdk/azure"
-import { streamText, generateText, ToolSet } from "ai"
+import { streamText, generateText, ToolSet, type ProviderMetadata } from "ai"
 
 import { azureOpenAiDefaultApiVersion, azureModels, azureDefaultModelInfo, type ModelInfo } from "@roo-code/types"
 
@@ -41,7 +41,7 @@ export class AzureHandler extends BaseProvider implements SingleCompletionHandle
 		// useDeploymentBasedUrls produces the universally compatible
 		// /deployments/{id}/{path} URL shape.
 		this.provider = createAzure({
-			baseURL: options.azureBaseUrl ?? "",
+			baseURL: options.azureBaseUrl || undefined,
 			apiKey: options.azureApiKey, // Optional — Azure supports managed identity / Entra ID auth
 			apiVersion: options.azureApiVersion ?? azureOpenAiDefaultApiVersion,
 			useDeploymentBasedUrls: true,
@@ -89,16 +89,14 @@ export class AzureHandler extends BaseProvider implements SingleCompletionHandle
 				reasoningTokens?: number
 			}
 		},
-		providerMetadata?: {
-			azure?: {
-				promptCacheHitTokens?: number
-				promptCacheMissTokens?: number
-			}
-		},
+		providerMetadata?: ProviderMetadata,
 	): ApiStreamUsageChunk {
 		// Extract cache metrics from Azure's providerMetadata if available
-		const cacheReadTokens = providerMetadata?.azure?.promptCacheHitTokens ?? usage.details?.cachedInputTokens
-		const cacheWriteTokens = providerMetadata?.azure?.promptCacheMissTokens
+		const azureMeta = providerMetadata?.azure as
+			| { promptCacheHitTokens?: number; promptCacheMissTokens?: number }
+			| undefined
+		const cacheReadTokens = azureMeta?.promptCacheHitTokens ?? usage.details?.cachedInputTokens
+		const cacheWriteTokens = azureMeta?.promptCacheMissTokens
 
 		return {
 			type: "usage",
@@ -165,7 +163,7 @@ export class AzureHandler extends BaseProvider implements SingleCompletionHandle
 			const usage = await result.usage
 			const providerMetadata = await result.providerMetadata
 			if (usage) {
-				yield this.processUsageMetrics(usage, providerMetadata as any)
+				yield this.processUsageMetrics(usage, providerMetadata)
 			}
 		} catch (error) {
 			// Handle AI SDK errors (AI_RetryError, AI_APICallError, etc.)

webview-ui/src/components/settings/constants.ts

@@ -52,7 +52,7 @@ export const PROVIDERS = [
 	{ value: "openrouter", label: "OpenRouter", proxy: false },
 	{ value: "deepinfra", label: "DeepInfra", proxy: false },
 	{ value: "anthropic", label: "Anthropic", proxy: false },
-	{ value: "azure", label: "Azure AI Foundry", proxy: false },
+	{ value: "azure", label: "Azure OpenAI", proxy: false },
 	{ value: "cerebras", label: "Cerebras", proxy: false },
 	{ value: "gemini", label: "Google Gemini", proxy: false },
 	{ value: "doubao", label: "Doubao", proxy: false },

webview-ui/src/components/settings/providers/Azure.tsx

@@ -6,6 +6,7 @@ import { type ProviderSettings, azureOpenAiDefaultApiVersion } from "@roo-code/t
 import { useAppTranslation } from "@src/i18n/TranslationContext"
 
 import { inputEventTransform } from "../transforms"
+import { parseAzureUrl } from "../utils/parseAzureUrl"
 
 type AzureProps = {
 	apiConfiguration: ProviderSettings
@@ -27,11 +28,29 @@ export const Azure = ({ apiConfiguration, setApiConfigurationField }: AzureProps
 		[setApiConfigurationField],
 	)
 
+	const handleBaseUrlInput = useCallback(
+		(event: unknown) => {
+			const rawValue = inputEventTransform(event)
+			const parsed = parseAzureUrl(rawValue)
+
+			if (parsed) {
+				setApiConfigurationField("azureBaseUrl", parsed.baseUrl)
+				setApiConfigurationField("azureDeploymentName", parsed.deploymentName)
+				if (parsed.apiVersion) {
+					setApiConfigurationField("azureApiVersion", parsed.apiVersion)
+				}
+			} else {
+				setApiConfigurationField("azureBaseUrl", rawValue)
+			}
+		},
+		[setApiConfigurationField],
+	)
+
 	return (
 		<>
 			<VSCodeTextField
 				value={apiConfiguration?.azureBaseUrl || ""}
-				onInput={handleInputChange("azureBaseUrl")}
+				onInput={handleBaseUrlInput}
 				placeholder={t("settings:placeholders.azureBaseUrl")}
 				className="w-full">
 				<label className="block font-medium mb-1">{t("settings:providers.azureBaseUrl")}</label>

webview-ui/src/components/settings/utils/__tests__/parseAzureUrl.spec.ts

@@ -0,0 +1,115 @@
+import { parseAzureUrl } from "../parseAzureUrl"
+
+describe("parseAzureUrl", () => {
+	it("parses a full openai.azure.com URL with api-version", () => {
+		const result = parseAzureUrl(
+			"https://my-resource.openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-10-21",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "gpt-4o",
+			apiVersion: "2024-10-21",
+		})
+	})
+
+	it("parses a cognitiveservices.azure.com URL", () => {
+		const result = parseAzureUrl(
+			"https://my-deployment.cognitiveservices.azure.com/openai/deployments/gpt-5.2/chat/completions?api-version=2024-05-01-preview",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-deployment.cognitiveservices.azure.com/openai",
+			deploymentName: "gpt-5.2",
+			apiVersion: "2024-05-01-preview",
+		})
+	})
+
+	it("parses a services.ai.azure.com URL", () => {
+		const result = parseAzureUrl(
+			"https://my-resource.services.ai.azure.com/openai/deployments/my-model/responses?api-version=2025-01-01",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.services.ai.azure.com/openai",
+			deploymentName: "my-model",
+			apiVersion: "2025-01-01",
+		})
+	})
+
+	it("handles URL without api-version query param", () => {
+		const result = parseAzureUrl("https://my-resource.openai.azure.com/openai/deployments/gpt-4o/chat/completions")
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "gpt-4o",
+		})
+	})
+
+	it("handles URL with trailing slash", () => {
+		const result = parseAzureUrl("https://my-resource.openai.azure.com/openai/deployments/gpt-4o/chat/completions/")
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "gpt-4o",
+		})
+	})
+
+	it("handles deployment name with dots", () => {
+		const result = parseAzureUrl(
+			"https://my-resource.openai.azure.com/openai/deployments/gpt-4.turbo.2024/chat/completions?api-version=2024-10-21",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "gpt-4.turbo.2024",
+			apiVersion: "2024-10-21",
+		})
+	})
+
+	it("handles URL with only /openai/deployments/{name} (no trailing path)", () => {
+		const result = parseAzureUrl(
+			"https://my-resource.openai.azure.com/openai/deployments/my-deploy?api-version=2024-10-21",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "my-deploy",
+			apiVersion: "2024-10-21",
+		})
+	})
+
+	it("returns null for a plain base URL (no /deployments/ path)", () => {
+		const result = parseAzureUrl("https://my-resource.openai.azure.com/openai")
+		expect(result).toBeNull()
+	})
+
+	it("returns null for a non-URL string", () => {
+		const result = parseAzureUrl("not-a-url")
+		expect(result).toBeNull()
+	})
+
+	it("returns null for an empty string", () => {
+		const result = parseAzureUrl("")
+		expect(result).toBeNull()
+	})
+
+	it("returns null for a URL without /openai/ prefix", () => {
+		const result = parseAzureUrl("https://my-resource.openai.azure.com/deployments/gpt-4o/chat/completions")
+		expect(result).toBeNull()
+	})
+
+	it("handles encoded deployment names", () => {
+		const result = parseAzureUrl(
+			"https://my-resource.openai.azure.com/openai/deployments/my%20deploy/chat/completions",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "my deploy",
+		})
+	})
+
+	it("handles additional query parameters besides api-version", () => {
+		const result = parseAzureUrl(
+			"https://my-resource.openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2024-10-21&extra=param",
+		)
+		expect(result).toEqual({
+			baseUrl: "https://my-resource.openai.azure.com/openai",
+			deploymentName: "gpt-4o",
+			apiVersion: "2024-10-21",
+		})
+	})
+})

webview-ui/src/components/settings/utils/parseAzureUrl.ts

@@ -0,0 +1,38 @@
+export interface ParsedAzureUrl {
+	/** e.g. "https://my-resource.cognitiveservices.azure.com/openai" */
+	baseUrl: string
+	/** e.g. "gpt-5.2" */
+	deploymentName: string
+	/** e.g. "2024-05-01-preview" */
+	apiVersion?: string
+}
+
+/**
+ * Parses a full Azure OpenAI URL into its components.
+ * Returns null if the URL doesn't match the expected pattern.
+ *
+ * Supported URL formats:
+ * - https://{resource}.openai.azure.com/openai/deployments/{deployment}/chat/completions?api-version={ver}
+ * - https://{resource}.cognitiveservices.azure.com/openai/deployments/{deployment}/responses?api-version={ver}
+ * - https://{resource}.services.ai.azure.com/openai/deployments/{deployment}/{anything}?api-version={ver}
+ */
+export function parseAzureUrl(input: string): ParsedAzureUrl | null {
+	let url: URL
+	try {
+		url = new URL(input)
+	} catch {
+		return null
+	}
+
+	// Match pathname: /openai/deployments/{name}/...
+	const match = url.pathname.match(/^(\/openai)\/deployments\/([^/]+)/)
+	if (!match) {
+		return null
+	}
+
+	const baseUrl = `${url.origin}${match[1]}`
+	const deploymentName = decodeURIComponent(match[2])
+	const apiVersion = url.searchParams.get("api-version") ?? undefined
+
+	return { baseUrl, deploymentName, ...(apiVersion !== undefined && { apiVersion }) }
+}

webview-ui/src/components/settings/utils/providerModelConfig.ts

@@ -33,7 +33,7 @@ export interface ProviderServiceConfig {
 export const PROVIDER_SERVICE_CONFIG: Partial<Record<ProviderName, ProviderServiceConfig>> = {
 	anthropic: { serviceName: "Anthropic", serviceUrl: "https://console.anthropic.com" },
 	azure: {
-		serviceName: "Azure AI Foundry",
+		serviceName: "Azure OpenAI",
 		serviceUrl: "https://azure.microsoft.com/en-us/products/ai-foundry/models/openai",
 	},
 	bedrock: { serviceName: "Amazon Bedrock", serviceUrl: "https://aws.amazon.com/bedrock" },

webview-ui/src/i18n/locales/en/settings.json

@@ -442,14 +442,14 @@
 		"getBasetenApiKey": "Get Baseten API Key",
 		"cerebrasApiKey": "Cerebras API Key",
 		"getCerebrasApiKey": "Get Cerebras API Key",
-		"azureBaseUrl": "Base URL",
-		"azureBaseUrlDescription": "Your Azure OpenAI endpoint URL. Found in the Azure portal under Keys & Endpoint.",
+		"azureBaseUrl": "Azure Endpoint",
+		"azureBaseUrlDescription": "Your Azure OpenAI endpoint. You can paste a full deployment URL (e.g., https://myresource.openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=...) and all fields below will auto-fill, or enter just the endpoint from the Azure portal (e.g., https://myresource.openai.azure.com/openai).",
 		"azureDeploymentName": "Azure Deployment Name",
-		"azureDeploymentNameDescription": "The name of your model deployment within the resource.",
+		"azureDeploymentNameDescription": "The name of your model deployment. This may differ from the model name — it's the name you chose when deploying the model in Azure.",
 		"azureApiKey": "Azure API Key",
 		"getAzureApiKey": "Get Azure OpenAI Access",
 		"azureApiVersion": "Azure API Version",
-		"azureApiVersionDescription": "The API version to use (e.g., '2024-10-21'). Leave empty for the default.",
+		"azureApiVersionDescription": "The API version to use. Leave empty to use the default (2025-04-01-preview).",
 		"chutesApiKey": "Chutes API Key",
 		"getChutesApiKey": "Get Chutes API Key",
 		"fireworksApiKey": "Fireworks API Key",