feat: add Google Vertex AI as new provider

env.example

@@ -1,6 +1,6 @@
 # AI Provider Configuration
 # AI_PROVIDER: Which provider to use
-# Options: bedrock, openai, anthropic, google, azure, ollama, openrouter, deepseek, siliconflow, gateway
+# Options: bedrock, openai, anthropic, google, vertexai, azure, ollama, openrouter, deepseek, siliconflow, gateway
 # Default: bedrock
 AI_PROVIDER=bedrock
 
@@ -38,7 +38,17 @@ AI_MODEL=global.anthropic.claude-sonnet-4-5-20250929-v1:0
 # GOOGLE_TOP_P=0.95                          # Optional: Nucleus sampling parameter
 # Note: Gemini 2.5/3 models automatically enable reasoning display (includeThoughts: true)
 # GOOGLE_THINKING_BUDGET=8192                # Optional: Gemini 2.5 thinking budget in tokens (for more/less thinking)
-# GOOGLE_THINKING_LEVEL=high                 # Optional: Gemini 3 thinking level (low/high)
+# GOOGLE_THINKING_LEVEL=high                 # Optional: Gemini 3 thinking level (minimal/low/medium/high)
+
+# Google Vertex AI Configuration (Enterprise GCP)
+# For enterprise users needing data residency, VPC Service Controls, or GCP integration
+# Uses GCP service account authentication instead of API keys
+# GOOGLE_VERTEX_PROJECT=your-gcp-project-id      # Required: GCP project ID
+# GOOGLE_VERTEX_LOCATION=us-central1              # Optional: defaults to us-central1
+# GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json  # Path to service account key file
+# Note: When running on GCP (Cloud Run, GKE, Compute Engine), uses Application Default Credentials automatically
+# GOOGLE_VERTEX_THINKING_BUDGET=8192              # Optional: Gemini 2.5 thinking budget in tokens (1024-100000)
+# GOOGLE_VERTEX_THINKING_LEVEL=high               # Optional: Gemini 3 thinking level (minimal/low/medium/high)
 
 # Azure OpenAI Configuration
 # Configure endpoint using ONE of these methods:

lib/ai-providers.ts

@@ -4,6 +4,7 @@ import { azure, createAzure } from "@ai-sdk/azure"
 import { createDeepSeek, deepseek } from "@ai-sdk/deepseek"
 import { createGateway, gateway } from "@ai-sdk/gateway"
 import { createGoogleGenerativeAI, google } from "@ai-sdk/google"
+import { createVertex } from "@ai-sdk/google-vertex"
 import { createOpenAI, openai } from "@ai-sdk/openai"
 import { fromNodeProviderChain } from "@aws-sdk/credential-providers"
 import { createOpenRouter } from "@openrouter/ai-sdk-provider"
@@ -38,6 +39,7 @@ const ALLOWED_CLIENT_PROVIDERS: ProviderName[] = [
     "openai",
     "anthropic",
     "google",
+    "vertexai",
     "azure",
     "bedrock",
     "openrouter",
@@ -95,7 +97,9 @@ function parseIntSafe(
  * - ANTHROPIC_THINKING_BUDGET_TOKENS: Anthropic thinking budget in tokens (1024-64000)
  * - ANTHROPIC_THINKING_TYPE: Anthropic thinking type (enabled)
  * - GOOGLE_THINKING_BUDGET: Google Gemini 2.5 thinking budget in tokens (1024-100000)
- * - GOOGLE_THINKING_LEVEL: Google Gemini 3 thinking level (low/high)
+ * - GOOGLE_THINKING_LEVEL: Google Gemini 3 thinking level (minimal/low/medium/high)
+ * - GOOGLE_VERTEX_THINKING_BUDGET: Vertex AI Gemini 2.5 thinking budget in tokens (1024-100000)
+ * - GOOGLE_VERTEX_THINKING_LEVEL: Vertex AI Gemini 3 thinking level (minimal/low/medium/high)
  * - AZURE_REASONING_EFFORT: Azure/OpenAI reasoning effort (low/medium/high)
  * - AZURE_REASONING_SUMMARY: Azure reasoning summary (none/brief/detailed)
  * - BEDROCK_REASONING_BUDGET_TOKENS: Bedrock Claude reasoning budget in tokens (1024-64000)
@@ -260,7 +264,39 @@ function buildProviderOptions(
             }
             break
         }
+        case "vertexai": {
+            // Google Vertex supports the same thinking config as standard Google provider
+            const thinkingBudget = parseIntSafe(
+                process.env.GOOGLE_VERTEX_THINKING_BUDGET,
+                "GOOGLE_VERTEX_THINKING_BUDGET",
+                1024,
+                100000,
+            )
+            const thinkingLevel = process.env.GOOGLE_VERTEX_THINKING_LEVEL
 
+            if (
+                modelId &&
+                (modelId.includes("gemini-2") ||
+                    modelId.includes("gemini-3") ||
+                    modelId.includes("gemini2") ||
+                    modelId.includes("gemini3"))
+            ) {
+                const thinkingConfig: Record<string, any> = {
+                    includeThoughts: true,
+                }
+                if (thinkingBudget) {
+                    thinkingConfig.thinkingBudget = thinkingBudget
+                } else if (thinkingLevel) {
+                    thinkingConfig.thinkingLevel = thinkingLevel as
+                        | "minimal"
+                        | "low"
+                        | "medium"
+                        | "high"
+                }
+                options.google = { thinkingConfig }
+            }
+            break
+        }
         case "azure": {
             const reasoningEffort = process.env.AZURE_REASONING_EFFORT
             const reasoningSummary = process.env.AZURE_REASONING_SUMMARY
@@ -362,6 +398,7 @@ const PROVIDER_ENV_VARS: Record<ProviderName, string | null> = {
     openai: "OPENAI_API_KEY",
     anthropic: "ANTHROPIC_API_KEY",
     google: "GOOGLE_GENERATIVE_AI_API_KEY",
+    vertexai: null, // Uses GOOGLE_APPLICATION_CREDENTIALS or IAM role
     azure: "AZURE_API_KEY",
     ollama: null, // No credentials needed for local Ollama
     openrouter: "OPENROUTER_API_KEY",
@@ -644,6 +681,26 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
             }
             break
         }
+        case "vertexai": {
+            // Google Vertex AI uses GCP service account authentication, not API keys
+            // Auth is handled via GOOGLE_APPLICATION_CREDENTIALS env var or GCP default credentials
+            const vertexProject = process.env.GOOGLE_VERTEX_PROJECT
+            const vertexLocation =
+                process.env.GOOGLE_VERTEX_LOCATION || "us-central1"
+
+            if (!vertexProject) {
+                throw new Error(
+                    "GOOGLE_VERTEX_PROJECT environment variable is required for vertexai provider.",
+                )
+            }
+
+            const vertexProvider = createVertex({
+                project: vertexProject,
+                location: vertexLocation,
+            })
+            model = vertexProvider(modelId)
+            break
+        }
 
         case "azure": {
             const apiKey = overrides?.apiKey || process.env.AZURE_API_KEY

lib/types/model-config.ts

@@ -4,6 +4,7 @@ export type ProviderName =
     | "openai"
     | "anthropic"
     | "google"
+    | "vertexai"
     | "azure"
     | "bedrock"
     | "ollama"
@@ -75,6 +76,7 @@ export const PROVIDER_INFO: Record<
         defaultBaseUrl: "https://api.anthropic.com/v1",
     },
     google: { label: "Google" },
+    vertexai: { label: "Google Vertex AI" },
     azure: { label: "Azure OpenAI" },
     bedrock: { label: "Amazon Bedrock" },
     ollama: {
@@ -157,6 +159,17 @@ export const SUGGESTED_MODELS: Partial<Record<ProviderName, string[]>> = {
         // Legacy
         "gemini-pro",
     ],
+    vertexai: [
+        // Gemini 2.5 series
+        "gemini-2.5-pro",
+        "gemini-2.5-flash",
+        // Gemini 2.0 series
+        "gemini-2.0-flash",
+        "gemini-2.0-flash-exp",
+        // Gemini 1.5 series
+        "gemini-1.5-pro",
+        "gemini-1.5-flash",
+    ],
     azure: ["gpt-4o", "gpt-4o-mini", "gpt-4-turbo", "gpt-4", "gpt-35-turbo"],
     bedrock: [
         // Anthropic Claude