diff --git a/package.json b/package.json
index 92618084d..cd59398ca 100644
--- a/package.json
+++ b/package.json
@@ -66,8 +66,8 @@
     "@e2b/code-interpreter": "^1.5.1",
     "@electron-toolkit/preload": "^3.0.2",
     "@electron-toolkit/utils": "^4.0.0",
-    "@google/genai": "^1.29.0",
     "node-pty": "^1.1.0-beta39",
+    "@google/genai": "^1.30.0",
     "@jxa/run": "^1.4.0",
     "@modelcontextprotocol/sdk": "^1.22.0",
     "axios": "^1.13.2",
diff --git a/src/main/presenter/configPresenter/modelCapabilities.ts b/src/main/presenter/configPresenter/modelCapabilities.ts
index 63607d5aa..47afdbb35 100644
--- a/src/main/presenter/configPresenter/modelCapabilities.ts
+++ b/src/main/presenter/configPresenter/modelCapabilities.ts
@@ -19,7 +19,8 @@ export class ModelCapabilities {
   private index: Map<string, Map<string, ProviderModel>> = new Map()
   private static readonly PROVIDER_ID_ALIASES: Record<string, string> = {
     dashscope: 'alibaba-cn',
-    gemini: 'google'
+    gemini: 'google',
+    vertex: 'google-vertex'
   }
 
   constructor() {
diff --git a/src/main/presenter/configPresenter/providers.ts b/src/main/presenter/configPresenter/providers.ts
index 68e112424..e0b17f86f 100644
--- a/src/main/presenter/configPresenter/providers.ts
+++ b/src/main/presenter/configPresenter/providers.ts
@@ -232,6 +232,21 @@ export const DEFAULT_PROVIDERS: LLM_PROVIDER_BASE[] = [
       defaultBaseUrl: 'https://generativelanguage.googleapis.com'
     }
   },
+  {
+    id: 'vertex',
+    name: 'Vertex AI',
+    apiType: 'vertex',
+    apiKey: '',
+    baseUrl: '',
+    enable: false,
+    websites: {
+      official: 'https://cloud.google.com/vertex-ai',
+      apiKey: 'https://console.cloud.google.com/apis/credentials',
+      docs: 'https://cloud.google.com/vertex-ai/generative-ai/docs',
+      models: 'https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/gemini',
+      defaultBaseUrl: 'https://aiplatform.googleapis.com'
+    }
+  },
   {
     id: 'anthropic',
     name: 'Anthropic',
diff --git a/src/main/presenter/llmProviderPresenter/managers/providerInstanceManager.ts b/src/main/presenter/llmProviderPresenter/managers/providerInstanceManager.ts
index 390f0af31..b624646bc 100644
--- a/src/main/presenter/llmProviderPresenter/managers/providerInstanceManager.ts
+++ b/src/main/presenter/llmProviderPresenter/managers/providerInstanceManager.ts
@@ -23,6 +23,7 @@ import { ZhipuProvider } from '../providers/zhipuProvider'
 import { LMStudioProvider } from '../providers/lmstudioProvider'
 import { OpenAIResponsesProvider } from '../providers/openAIResponsesProvider'
 import { CherryInProvider } from '../providers/cherryInProvider'
+import { VertexProvider } from '../providers/vertexProvider'
 import { OpenRouterProvider } from '../providers/openRouterProvider'
 import { MinimaxProvider } from '../providers/minimaxProvider'
 import { AihubmixProvider } from '../providers/aihubmixProvider'
@@ -77,6 +78,7 @@ export class ProviderInstanceManager {
       ['dashscope', DashscopeProvider],
       ['gemini', GeminiProvider],
       ['zhipu', ZhipuProvider],
+      ['vertex', VertexProvider],
       ['github', GithubProvider],
       ['github-copilot', GithubCopilotProvider],
       ['ollama', OllamaProvider],
@@ -106,6 +108,7 @@ export class ProviderInstanceManager {
       ['dashscope', DashscopeProvider],
       ['ppio', PPIOProvider],
       ['gemini', GeminiProvider],
+      ['vertex', VertexProvider],
       ['zhipu', ZhipuProvider],
       ['github', GithubProvider],
       ['github-copilot', GithubCopilotProvider],
diff --git a/src/main/presenter/llmProviderPresenter/providers/vertexProvider.ts b/src/main/presenter/llmProviderPresenter/providers/vertexProvider.ts
new file mode 100644
index 000000000..e09670735
--- /dev/null
+++ b/src/main/presenter/llmProviderPresenter/providers/vertexProvider.ts
@@ -0,0 +1,1193 @@
+import { presenter } from '@/presenter'
+import {
+  Content,
+  FunctionCallingConfigMode,
+  GenerateContentParameters,
+  GenerateContentResponseUsageMetadata,
+  GoogleGenAI,
+  HarmBlockThreshold,
+  HarmCategory,
+  Modality,
+  Part,
+  SafetySetting,
+  Tool,
+  GoogleSearch,
+  GenerateContentConfig
+} from '@google/genai'
+import { ModelType } from '@shared/model'
+import {
+  ChatMessage,
+  IConfigPresenter,
+  LLM_PROVIDER,
+  LLMCoreStreamEvent,
+  LLMResponse,
+  MCPToolDefinition,
+  MODEL_META,
+  ModelConfig
+} from '@shared/presenter'
+import { VERTEX_PROVIDER } from '@shared/types/presenters/llmprovider.presenter'
+import { createStreamEvent } from '@shared/types/core/llm-events'
+import { BaseLLMProvider, SUMMARY_TITLES_PROMPT } from '../baseProvider'
+import { modelCapabilities } from '../../configPresenter/modelCapabilities'
+import { eventBus, SendTarget } from '@/eventbus'
+import { CONFIG_EVENTS } from '@/events'
+
+// Mapping from simple keys to API HarmCategory constants
+const keyToHarmCategoryMap: Record<string, HarmCategory> = {
+  harassment: HarmCategory.HARM_CATEGORY_HARASSMENT,
+  hateSpeech: HarmCategory.HARM_CATEGORY_HATE_SPEECH,
+  sexuallyExplicit: HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
+  dangerousContent: HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT
+}
+
+// Value mapping from config storage to API HarmBlockThreshold constants
+// Assuming config stores 'BLOCK_NONE', 'BLOCK_LOW_AND_ABOVE', etc. directly
+const valueToHarmBlockThresholdMap: Record<string, HarmBlockThreshold> = {
+  BLOCK_NONE: HarmBlockThreshold.BLOCK_NONE,
+  BLOCK_LOW_AND_ABOVE: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE,
+  BLOCK_MEDIUM_AND_ABOVE: HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE,
+  BLOCK_ONLY_HIGH: HarmBlockThreshold.BLOCK_ONLY_HIGH,
+  HARM_BLOCK_THRESHOLD_UNSPECIFIED: HarmBlockThreshold.HARM_BLOCK_THRESHOLD_UNSPECIFIED
+}
+const safetySettingKeys = Object.keys(keyToHarmCategoryMap)
+
+export class VertexProvider extends BaseLLMProvider {
+  private genAI: GoogleGenAI
+
+  constructor(provider: LLM_PROVIDER, configPresenter: IConfigPresenter) {
+    super(provider, configPresenter)
+    this.genAI = this.createGenAIClient()
+    this.init()
+  }
+
+  public onProxyResolved(): void {
+    this.genAI = this.createGenAIClient()
+    this.init()
+  }
+
+  private get vertexProvider(): VERTEX_PROVIDER {
+    return this.provider as VERTEX_PROVIDER
+  }
+
+  private resolveProjectId(): string | undefined {
+    return this.vertexProvider.projectId || process.env.GOOGLE_CLOUD_PROJECT
+  }
+
+  private resolveLocation(): string | undefined {
+    return this.vertexProvider.location || process.env.GOOGLE_CLOUD_LOCATION
+  }
+
+  private buildGoogleAuthOptions() {
+    const privateKey = this.vertexProvider.accountPrivateKey
+    const clientEmail = this.vertexProvider.accountClientEmail
+    if (privateKey && clientEmail) {
+      return {
+        projectId: this.resolveProjectId(),
+        credentials: {
+          client_email: clientEmail,
+          private_key: privateKey.replace(/\\n/g, '\n')
+        },
+        scopes: ['https://www.googleapis.com/auth/cloud-platform']
+      }
+    }
+    return undefined
+  }
+
+  private buildBaseUrl(): string {
+    const customBaseUrl = this.vertexProvider.baseUrl?.trim()
+    if (customBaseUrl) return customBaseUrl
+
+    const location = this.resolveLocation() || 'us-central1'
+    if (this.vertexProvider.endpointMode === 'express') {
+      return 'https://aiplatform.googleapis.com/'
+    }
+    return `https://${location}-aiplatform.googleapis.com/`
+  }
+
+  private getApiVersion(): 'v1' | 'v1beta1' {
+    return (this.vertexProvider.apiVersion as 'v1' | 'v1beta1') || 'v1'
+  }
+
+  private createGenAIClient(): GoogleGenAI {
+    const project = this.resolveProjectId()
+    const location = this.resolveLocation()
+    const apiVersion = this.getApiVersion()
+
+    return new GoogleGenAI({
+      vertexai: true,
+      project,
+      location,
+      apiVersion,
+      googleAuthOptions: this.buildGoogleAuthOptions(),
+      httpOptions: {
+        baseUrl: this.buildBaseUrl(),
+        apiVersion
+      }
+    })
+  }
+
+  // 确保带有 Vertex 格式的模型路径
+  private ensureVertexModelName(modelId: string): string {
+    if (!modelId) return modelId
+    if (modelId.startsWith('projects/')) return modelId
+    const normalized = modelId
+      .replace(/^models\//i, '')
+      .replace(/^publishers\/google\/models\//i, '')
+    return `publishers/google/models/${normalized}`
+  }
+
+  // Implement abstract method fetchProviderModels from BaseLLMProvider
+  protected async fetchProviderModels(): Promise<MODEL_META[]> {
+    try {
+      const modelsResponse = await this.genAI.models.list()
+      // console.log('gemini models response:', modelsResponse)
+
+      // 将 pager 转换为数组
+      const models: any[] = []
+      for await (const model of modelsResponse) {
+        models.push(model)
+      }
+
+      if (models.length === 0) {
+        console.warn('No models found in Vertex AI response, using Provider DB models')
+        const dbModels = this.configPresenter.getDbProviderModels(this.provider.id).map((m) => ({
+          id: m.id,
+          name: m.name,
+          group: m.group || 'default',
+          providerId: this.provider.id,
+          isCustom: false,
+          contextLength: m.contextLength,
+          maxTokens: m.maxTokens,
+          vision: m.vision || false,
+          functionCall: m.functionCall || false,
+          reasoning: m.reasoning || false,
+          ...(m.type ? { type: m.type } : {})
+        }))
+        return dbModels
+      }
+
+      // 映射 API 返回的模型数据（能力统一读 Provider DB）
+      const normalizeModelId = (mid: string): string =>
+        String(mid || '')
+          .replace(/^models\//i, '')
+          .replace(/^publishers\/google\/models\//i, '')
+      const apiModels: MODEL_META[] = models
+        .filter((model: any) => {
+          const name = String(model.name || '').toLowerCase()
+          return (
+            !name.includes('embedding') &&
+            !name.includes('aqa') &&
+            !name.includes('text-embedding') &&
+            !name.includes('gemma-3n-e4b-it')
+          )
+        })
+        .map((model: any) => {
+          const apiModelId: string = model.name
+          const displayName: string = model.displayName || apiModelId
+
+          const normalizedId = normalizeModelId(apiModelId)
+
+          const vision = modelCapabilities.supportsVision(this.provider.id, normalizedId)
+          const functionCall = modelCapabilities.supportsToolCall(this.provider.id, normalizedId)
+          const reasoning = modelCapabilities.supportsReasoning(this.provider.id, normalizedId)
+          const isImageOutput = modelCapabilities.supportsImageOutput(
+            this.provider.id,
+            normalizedId
+          )
+          const modelType = isImageOutput ? ModelType.ImageGeneration : ModelType.Chat
+
+          let group = 'default'
+          if (/\b(exp|preview)\b/i.test(apiModelId)) group = 'experimental'
+          else if (/\bgemma\b/i.test(apiModelId)) group = 'gemma'
+
+          return {
+            id: apiModelId,
+            name: displayName,
+            group,
+            providerId: this.provider.id,
+            isCustom: false,
+            contextLength: model.inputTokenLimit,
+            maxTokens: model.outputTokenLimit,
+            vision,
+            functionCall,
+            reasoning,
+            ...(modelType !== ModelType.Chat && { type: modelType })
+          } as MODEL_META
+        })
+
+      // console.log('Mapped Vertex models:', apiModels)
+      return apiModels
+    } catch (error) {
+      console.warn('Failed to fetch models from Vertex AI:', error)
+      // If API call fails, fallback to Provider DB mapping
+      const dbModels = this.configPresenter.getDbProviderModels(this.provider.id).map((m) => ({
+        id: m.id,
+        name: m.name,
+        group: m.group || 'default',
+        providerId: this.provider.id,
+        isCustom: false,
+        contextLength: m.contextLength,
+        maxTokens: m.maxTokens,
+        vision: m.vision || false,
+        functionCall: m.functionCall || false,
+        reasoning: m.reasoning || false,
+        ...(m.type ? { type: m.type } : {})
+      }))
+      return dbModels
+    }
+  }
+
+  // Implement summaryTitles abstract method from BaseLLMProvider
+  public async summaryTitles(
+    messages: { role: 'system' | 'user' | 'assistant'; content: string }[],
+    modelId: string
+  ): Promise<string> {
+    console.log('vertex summary check, ignore modelId', modelId)
+    // Use Vertex AI to generate conversation titles
+    try {
+      const conversationText = messages.map((m) => `${m.role}: ${m.content}`).join('\n')
+      const prompt = `${SUMMARY_TITLES_PROMPT}\n\n${conversationText}`
+
+      const result = await this.genAI.models.generateContent({
+        model: this.ensureVertexModelName(modelId),
+        contents: [{ role: 'user', parts: [{ text: prompt }] }],
+        config: this.getGenerateContentConfig(0.4, undefined, modelId, false)
+      })
+
+      return result.text?.trim() || 'New Conversation'
+    } catch (error) {
+      console.error('Failed to generate conversation title:', error)
+      return 'New Conversation'
+    }
+  }
+
+  // Override fetchModels method since Vertex AI will reuse the cached model list
+  async fetchModels(): Promise<MODEL_META[]> {
+    // Vertex AI will reuse the cached model list
+    return this.models
+  }
+
+  // Override check method to use the first default model for testing
+  async check(): Promise<{ isOk: boolean; errorMsg: string | null }> {
+    try {
+      const projectId = this.resolveProjectId()
+      const location = this.resolveLocation()
+      if (!projectId || !location) {
+        return { isOk: false, errorMsg: 'projectId and location are required for Vertex AI' }
+      }
+      if (this.vertexProvider.accountPrivateKey && !this.vertexProvider.accountClientEmail) {
+        return {
+          isOk: false,
+          errorMsg: 'accountClientEmail is required when accountPrivateKey is provided'
+        }
+      }
+      this.genAI = this.createGenAIClient()
+
+      // Use the first model for simple testing
+      const testModelId =
+        this.models.find((m) => m.type !== ModelType.ImageGeneration)?.id ||
+        this.models[0]?.id ||
+        'gemini-1.5-flash-001'
+
+      const result = await this.genAI.models.generateContent({
+        model: this.ensureVertexModelName(testModelId),
+        contents: [{ role: 'user', parts: [{ text: 'Hello from Vertex AI' }] }]
+      })
+      return {
+        isOk: Boolean(result && (result.text || result.candidates?.length)),
+        errorMsg: null
+      }
+    } catch (error) {
+      console.error('Vertex provider check failed:', this.provider.name, error)
+      return { isOk: false, errorMsg: error instanceof Error ? error.message : String(error) }
+    }
+  }
+
+  protected async init() {
+    if (this.provider.enable) {
+      try {
+        const projectId = this.resolveProjectId()
+        const location = this.resolveLocation()
+        if (!projectId || !location) {
+          console.warn('Vertex provider missing projectId or location, skip initialization')
+          return
+        }
+        this.genAI = this.createGenAIClient()
+        this.isInitialized = true
+        // Use API to get model list, fallback to static list if failed
+        this.models = await this.fetchProviderModels()
+        await this.autoEnableModelsIfNeeded()
+        // Vertex AI is relatively slow, special compensation
+        eventBus.sendToRenderer(
+          CONFIG_EVENTS.MODEL_LIST_CHANGED,
+          SendTarget.ALL_WINDOWS,
+          this.provider.id
+        )
+        console.info('Provider initialized successfully:', this.provider.name)
+      } catch (error) {
+        console.warn('Provider initialization failed:', this.provider.name, error)
+      }
+    }
+  }
+
+  /**
+   * 重写 autoEnableModelsIfNeeded 方法
+   * 不自动启用模型，交由用户手动选择。
+   */
+  protected async autoEnableModelsIfNeeded() {
+    if (!this.models || this.models.length === 0) return
+    const providerId = this.provider.id
+
+    // 检查是否有自定义模型
+    const customModels = this.configPresenter.getCustomModels(providerId)
+    if (customModels && customModels.length > 0) return
+
+    // 检查是否有任何模型的状态被手动修改过
+    const hasManuallyModifiedModels = this.models.some((model) =>
+      this.configPresenter.getModelStatus(providerId, model.id)
+    )
+    if (hasManuallyModifiedModels) return
+
+    // 检查是否有任何已启用的模型
+    const hasEnabledModels = this.models.some((model) =>
+      this.configPresenter.getModelStatus(providerId, model.id)
+    )
+
+    // 不再自动启用模型，让用户手动选择启用需要的模型
+    if (!hasEnabledModels) {
+      console.info(
+        `Provider ${this.provider.name} models loaded, please manually enable the models you need`
+      )
+    }
+  }
+
+  // Helper function to get and format safety settings
+  private async getFormattedSafetySettings(): Promise<SafetySetting[] | undefined> {
+    const safetySettings: SafetySetting[] = []
+
+    for (const key of safetySettingKeys) {
+      try {
+        // Use configPresenter to get the setting value for the 'gemini' provider
+        // Assuming getSetting returns the string value like 'BLOCK_MEDIUM_AND_ABOVE'
+        const settingValue =
+          (await this.configPresenter.getSetting<string>(
+            `geminiSafety_${key}` // Match the key used in settings store
+          )) || 'HARM_BLOCK_THRESHOLD_UNSPECIFIED' // Default if not set
+
+        const threshold = valueToHarmBlockThresholdMap[settingValue]
+        const category = keyToHarmCategoryMap[key]
+
+        // Only add if threshold is defined, category is defined, and threshold is not BLOCK_NONE
+        if (
+          threshold &&
+          category &&
+          threshold !== 'BLOCK_NONE' &&
+          threshold !== 'HARM_BLOCK_THRESHOLD_UNSPECIFIED'
+        ) {
+          safetySettings.push({ category, threshold })
+        }
+      } catch (error) {
+        console.warn(`Failed to retrieve or map safety setting for ${key}:`, error)
+      }
+    }
+
+    return safetySettings.length > 0 ? safetySettings : undefined
+  }
+
+  // 判断模型是否支持 thinkingBudget
+  private supportsThinkingBudget(modelId: string): boolean {
+    const normalized = modelId.replace(/^models\//i, '')
+    const range = modelCapabilities.getThinkingBudgetRange(this.provider.id, normalized)
+    return (
+      typeof range.default === 'number' ||
+      typeof range.min === 'number' ||
+      typeof range.max === 'number'
+    )
+  }
+
+  // 获取生成配置，不再创建模型实例
+  private getGenerateContentConfig(
+    temperature?: number,
+    maxTokens?: number,
+    modelId?: string,
+    reasoning?: boolean,
+    thinkingBudget?: number
+  ): GenerateContentConfig {
+    const config: GenerateContentConfig = {
+      temperature,
+      maxOutputTokens: maxTokens,
+      topP: 1 // topP默认为1.0
+    }
+
+    // 从当前模型列表中查找指定的模型
+    if (modelId && this.models) {
+      const model = this.models.find((m) => m.id === modelId)
+      if (model && model.type === ModelType.ImageGeneration) {
+        config.responseModalities = [Modality.TEXT, Modality.IMAGE]
+      }
+    }
+
+    // 正确配置思考功能
+    if (reasoning) {
+      config.thinkingConfig = {
+        includeThoughts: true
+      }
+
+      // 仅对支持 thinkingBudget 的 Gemini 2.5 系列模型添加 thinkingBudget 参数
+      if (modelId && this.supportsThinkingBudget(modelId) && thinkingBudget !== undefined) {
+        config.thinkingConfig.thinkingBudget = thinkingBudget
+      }
+    }
+
+    return config
+  }
+
+  // 将 ChatMessage 转换为 Gemini 格式的消息
+  private formatVertexMessages(messages: ChatMessage[]): {
+    systemInstruction: string
+    contents: Content[]
+  } {
+    // 提取系统消息
+    const systemMessages = messages.filter((msg) => msg.role === 'system')
+    let systemContent = ''
+    if (systemMessages.length > 0) {
+      systemContent = systemMessages.map((msg) => msg.content).join('\n')
+    }
+
+    // 创建Gemini内容数组
+    const formattedContents: Content[] = []
+
+    // 处理非系统消息
+    const nonSystemMessages = messages.filter((msg) => msg.role !== 'system')
+    for (let i = 0; i < nonSystemMessages.length; i++) {
+      const message = nonSystemMessages[i]
+
+      // 检查是否是带有tool_calls的assistant消息
+      if (message.role === 'assistant' && 'tool_calls' in message) {
+        // 处理tool_calls消息
+        for (const toolCall of message.tool_calls || []) {
+          // 添加模型发出的函数调用
+          formattedContents.push({
+            role: 'model',
+            parts: [
+              {
+                functionCall: {
+                  name: toolCall.function.name,
+                  args: JSON.parse(toolCall.function.arguments || '{}')
+                }
+              }
+            ]
+          })
+
+          // 查找对应的工具响应消息
+          const nextMessage = i + 1 < nonSystemMessages.length ? nonSystemMessages[i + 1] : null
+          if (
+            nextMessage &&
+            nextMessage.role === 'tool' &&
+            'tool_call_id' in nextMessage &&
+            nextMessage.tool_call_id === toolCall.id
+          ) {
+            // 添加用户角色的函数响应
+            formattedContents.push({
+              role: 'user',
+              parts: [
+                {
+                  functionResponse: {
+                    name: toolCall.function.name,
+                    response: {
+                      result:
+                        typeof nextMessage.content === 'string'
+                          ? nextMessage.content
+                          : JSON.stringify(nextMessage.content)
+                    }
+                  }
+                }
+              ]
+            })
+
+            // 跳过下一条消息，因为已经处理过了
+            i++
+          }
+        }
+        continue
+      }
+
+      // 为每条消息创建parts数组
+      const parts: Part[] = []
+
+      // 检查消息是否包含工具调用或工具响应
+      if (message.role === 'tool' && Array.isArray(message.content)) {
+        // 处理工具消息
+        for (const part of message.content) {
+          // @ts-ignore - 处理类型兼容性
+          if (part.type === 'function_call' && part.function_call) {
+            // 处理函数调用
+            parts.push({
+              // @ts-ignore - 处理类型兼容性
+              functionCall: {
+                // @ts-ignore - 处理类型兼容性
+                name: part.function_call.name || '',
+                // @ts-ignore - 处理类型兼容性
+                args: part.function_call.arguments ? JSON.parse(part.function_call.arguments) : {}
+              }
+            })
+            // @ts-ignore - 处理类型兼容性
+          } else if (part.type === 'function_response') {
+            // 处理函数响应
+            // @ts-ignore - 处理类型兼容性
+            parts.push({ text: part.function_response || '' })
+          }
+        }
+      } else if (typeof message.content === 'string') {
+        // 处理消息内容 - 可能是字符串或包含图片的数组
+        // 处理纯文本消息
+        // 只添加非空文本
+        if (message.content.trim() !== '') {
+          parts.push({ text: message.content })
+        }
+      } else if (Array.isArray(message.content)) {
+        // 处理多模态消息（带图片等）
+        for (const part of message.content) {
+          if (part.type === 'text') {
+            // 只添加非空文本
+            if (part.text && part.text.trim() !== '') {
+              parts.push({ text: part.text })
+            }
+          } else if (part.type === 'image_url' && part.image_url) {
+            // 处理图片（假设是 base64 格式）
+            const matches = part.image_url.url.match(/^data:([^;]+);base64,(.+)$/)
+            if (matches && matches.length === 3) {
+              const mimeType = matches[1]
+              const base64Data = matches[2]
+              parts.push({
+                inlineData: {
+                  data: base64Data,
+                  mimeType: mimeType
+                }
+              })
+            }
+          }
+        }
+      }
+
+      // 只有当parts不为空时，才添加到formattedContents中
+      if (parts.length > 0) {
+        // 将消息角色转换为Gemini支持的角色
+        let role: 'user' | 'model' = 'user'
+        if (message.role === 'assistant') {
+          role = 'model'
+        } else if (message.role === 'tool') {
+          // 工具消息作为用户消息处理
+          role = 'user'
+        }
+
+        formattedContents.push({
+          role: role,
+          parts: parts
+        })
+      }
+    }
+
+    return { systemInstruction: systemContent, contents: formattedContents }
+  }
+
+  // 处理 Vertex API 响应，支持新旧格式的思考内容
+  private processVertexResponse(result: any): LLMResponse {
+    const resultResp: LLMResponse = {
+      content: ''
+    }
+
+    let textContent = ''
+    let thoughtContent = ''
+
+    // 检查是否有候选响应和 parts
+    if (result.candidates && result.candidates[0]?.content?.parts) {
+      for (const part of result.candidates[0].content.parts) {
+        // 检查是否是思考内容 (新格式)
+        if ((part as any).thought === true && part.text) {
+          thoughtContent += part.text
+        } else if (part.text) {
+          textContent += part.text
+        }
+      }
+    } else {
+      // 回退到使用 result.text
+      textContent = result.text || ''
+    }
+
+    // 如果没有检测到新格式的思考内容，检查旧格式的 <think> 标签
+    if (!thoughtContent && textContent.includes('<think>')) {
+      const thinkStart = textContent.indexOf('<think>')
+      const thinkEnd = textContent.indexOf('</think>')
+
+      if (thinkEnd > thinkStart) {
+        // 提取reasoning_content
+        thoughtContent = textContent.substring(thinkStart + 7, thinkEnd).trim()
+
+        // 合并<think>前后的普通内容
+        const beforeThink = textContent.substring(0, thinkStart).trim()
+        const afterThink = textContent.substring(thinkEnd + 8).trim()
+        textContent = [beforeThink, afterThink].filter(Boolean).join('\n')
+      }
+    }
+
+    resultResp.content = textContent
+    if (thoughtContent) {
+      resultResp.reasoning_content = thoughtContent
+    }
+
+    return resultResp
+  }
+
+  // 实现抽象方法
+  async completions(
+    messages: { role: 'system' | 'user' | 'assistant'; content: string }[],
+    modelId: string,
+    temperature?: number,
+    maxTokens?: number
+  ): Promise<LLMResponse> {
+    try {
+      if (!this.genAI) {
+        throw new Error('Google Generative AI client is not initialized')
+      }
+
+      const { systemInstruction, contents } = this.formatVertexMessages(messages)
+
+      // 创建 GenerateContentConfig
+      const generateContentConfig: GenerateContentConfig = this.getGenerateContentConfig(
+        temperature ?? 0.7,
+        maxTokens,
+        modelId,
+        false // completions 方法中不处理 reasoning
+      )
+
+      if (systemInstruction) {
+        generateContentConfig.systemInstruction = systemInstruction
+      }
+
+      // 一次性创建 requestParams
+      const requestParams: GenerateContentParameters = {
+        model: modelId,
+        contents,
+        config: generateContentConfig
+      }
+
+      const result = await this.genAI.models.generateContent({
+        ...requestParams,
+        model: this.ensureVertexModelName(requestParams.model as string)
+      })
+
+      const resultResp: LLMResponse = {
+        content: ''
+      }
+
+      // 尝试获取tokens信息 - 使用新SDK的usageMetadata结构
+      try {
+        if (result.usageMetadata) {
+          const usage = result.usageMetadata
+          resultResp.totalUsage = {
+            prompt_tokens: usage.promptTokenCount || 0,
+            completion_tokens: usage.candidatesTokenCount || 0,
+            total_tokens: usage.totalTokenCount || 0
+          }
+        } else {
+          // 估算token数量 - 简单方法，可以根据实际需要调整
+          const promptText = messages.map((m) => m.content).join(' ')
+          const responseText = result.text || ''
+
+          // 简单估算: 英文约1个token/4个字符，中文约1个token/1.5个字符
+          const estimateTokens = (text: string): number => {
+            const chineseCharCount = (text.match(/[\u4e00-\u9fa5]/g) || []).length
+            const otherCharCount = text.length - chineseCharCount
+            return Math.ceil(chineseCharCount / 1.5 + otherCharCount / 4)
+          }
+
+          const promptTokens = estimateTokens(promptText)
+          const completionTokens = estimateTokens(responseText)
+
+          resultResp.totalUsage = {
+            prompt_tokens: promptTokens,
+            completion_tokens: completionTokens,
+            total_tokens: promptTokens + completionTokens
+          }
+        }
+      } catch (e) {
+        console.warn('Failed to estimate token count for Vertex response', e)
+      }
+
+      // 处理响应内容，支持新格式的思考内容
+      let textContent = ''
+      let thoughtContent = ''
+
+      // 检查是否有候选响应和 parts
+      if (result.candidates && result.candidates[0]?.content?.parts) {
+        for (const part of result.candidates[0].content.parts) {
+          // 检查是否是思考内容 (新格式)
+          if ((part as any).thought === true && part.text) {
+            thoughtContent += part.text
+          } else if (part.text) {
+            textContent += part.text
+          }
+        }
+      } else {
+        // 回退到使用 result.text
+        textContent = result.text || ''
+      }
+
+      // 如果没有检测到新格式的思考内容，检查旧格式的 <think> 标签
+      if (!thoughtContent && textContent.includes('<think>')) {
+        const thinkStart = textContent.indexOf('<think>')
+        const thinkEnd = textContent.indexOf('</think>')
+
+        if (thinkEnd > thinkStart) {
+          // 提取reasoning_content
+          thoughtContent = textContent.substring(thinkStart + 7, thinkEnd).trim()
+
+          // 合并<think>前后的普通内容
+          const beforeThink = textContent.substring(0, thinkStart).trim()
+          const afterThink = textContent.substring(thinkEnd + 8).trim()
+          textContent = [beforeThink, afterThink].filter(Boolean).join('\n')
+        }
+      }
+
+      resultResp.content = textContent
+      if (thoughtContent) {
+        resultResp.reasoning_content = thoughtContent
+      }
+
+      return resultResp
+    } catch (error) {
+      console.error('Vertex completions error:', error)
+      throw error
+    }
+  }
+
+  async summaries(
+    text: string,
+    modelId: string,
+    temperature?: number,
+    maxTokens?: number
+  ): Promise<LLMResponse> {
+    if (!this.isInitialized) {
+      throw new Error('Provider not initialized')
+    }
+
+    if (!modelId) {
+      throw new Error('Model ID is required')
+    }
+
+    try {
+      const prompt = `Please generate a concise summary for the following content:\n\n${text}`
+
+      const result = await this.genAI.models.generateContent({
+        model: this.ensureVertexModelName(modelId),
+        contents: [{ role: 'user', parts: [{ text: prompt }] }],
+        config: this.getGenerateContentConfig(temperature, maxTokens, modelId, false)
+      })
+
+      return this.processVertexResponse(result)
+    } catch (error) {
+      console.error('Vertex summaries error:', error)
+      throw error
+    }
+  }
+
+  async generateText(
+    prompt: string,
+    modelId: string,
+    temperature?: number,
+    maxTokens?: number
+  ): Promise<LLMResponse> {
+    if (!this.isInitialized) {
+      throw new Error('Provider not initialized')
+    }
+
+    if (!modelId) {
+      throw new Error('Model ID is required')
+    }
+
+    try {
+      const result = await this.genAI.models.generateContent({
+        model: this.ensureVertexModelName(modelId),
+        contents: [{ role: 'user', parts: [{ text: prompt }] }],
+        config: this.getGenerateContentConfig(temperature, maxTokens, modelId, false)
+      })
+
+      return this.processVertexResponse(result)
+    } catch (error) {
+      console.error('Vertex generateText error:', error)
+      throw error
+    }
+  }
+
+  async suggestions(
+    context: string,
+    modelId: string,
+    temperature?: number,
+    maxTokens?: number
+  ): Promise<string[]> {
+    if (!this.isInitialized) {
+      throw new Error('Provider not initialized')
+    }
+
+    if (!modelId) {
+      throw new Error('Model ID is required')
+    }
+
+    try {
+      const prompt = `Based on the following context, please provide up to 5 reasonable suggestion options, each not exceeding 100 characters. Please return in JSON array format without other explanations:\n\n${context}`
+
+      const result = await this.genAI.models.generateContent({
+        model: this.ensureVertexModelName(modelId),
+        contents: [{ role: 'user', parts: [{ text: prompt }] }],
+        config: this.getGenerateContentConfig(temperature, maxTokens, modelId, false)
+      })
+
+      const responseText = result.text || ''
+
+      // 尝试从响应中解析出JSON数组
+      try {
+        const cleanedText = responseText.replace(/```json|```/g, '').trim()
+        const suggestions = JSON.parse(cleanedText)
+        if (Array.isArray(suggestions)) {
+          return suggestions.map((item) => item.toString())
+        }
+      } catch (parseError) {
+        console.error('Vertex suggestions parseError:', parseError)
+        // 如果解析失败，尝试分行处理
+        const lines = responseText
+          .split('\n')
+          .map((line) => line.trim())
+          .filter((line) => line && !line.startsWith('```') && !line.includes(':'))
+          .map((line) => line.replace(/^[0-9]+\.\s*/, '').replace(/^-\s*/, ''))
+
+        if (lines.length > 0) {
+          return lines.slice(0, 5)
+        }
+      }
+
+      // If all fail, return a default prompt
+      return ['Unable to generate suggestions']
+    } catch (error) {
+      console.error('Vertex suggestions error:', error)
+      return ['Error occurred, unable to get suggestions']
+    }
+  }
+  /**
+   * 核心流式处理方法
+   * 实现BaseLLMProvider中的抽象方法
+   */
+  async *coreStream(
+    messages: ChatMessage[],
+    modelId: string,
+    modelConfig: ModelConfig,
+    temperature: number,
+    maxTokens: number,
+    mcpTools: MCPToolDefinition[]
+  ): AsyncGenerator<LLMCoreStreamEvent> {
+    if (!this.isInitialized) throw new Error('Provider not initialized')
+    if (!modelId) throw new Error('Model ID is required')
+    console.log('modelConfig', modelConfig, modelId)
+
+    // 检查是否是图片生成模型
+    const isImageGenerationModel = modelConfig?.type === ModelType.ImageGeneration
+
+    // 如果是图片生成模型，使用特殊处理
+    if (isImageGenerationModel) {
+      yield* this.handleImageGenerationStream(messages, modelId, temperature, maxTokens)
+      return
+    }
+
+    const safetySettings = await this.getFormattedSafetySettings()
+    console.log('safetySettings', safetySettings)
+
+    // 添加Gemini工具调用
+    let geminiTools: Tool[] = []
+
+    // 注意：googleSearch内置工具与外部工具是互斥的
+    if (modelConfig.enableSearch) {
+      geminiTools.push({ googleSearch: {} as GoogleSearch })
+    } else {
+      if (mcpTools.length > 0)
+        geminiTools = await presenter.mcpPresenter.mcpToolsToGeminiTools(mcpTools, this.provider.id)
+    }
+
+    // 格式化消息为Gemini格式
+    const formattedParts = this.formatVertexMessages(messages)
+
+    // 1. 获取基础 config
+    const generateContentConfig: GenerateContentConfig = this.getGenerateContentConfig(
+      temperature,
+      maxTokens,
+      modelId,
+      modelConfig.reasoning,
+      modelConfig.thinkingBudget
+    )
+
+    // 2. 在本地变量上添加其他属性
+    if (formattedParts.systemInstruction) {
+      generateContentConfig.systemInstruction = formattedParts.systemInstruction
+    }
+
+    if (geminiTools.length > 0) {
+      generateContentConfig.tools = geminiTools
+      // 仅当存在 functionDeclarations 时才配置 functionCallingConfig
+      const hasFunctionDeclarations = geminiTools.some((t: any) => {
+        const fns = t?.functionDeclarations
+        return Array.isArray(fns) && fns.length > 0
+      })
+      if (hasFunctionDeclarations) {
+        generateContentConfig.toolConfig = {
+          functionCallingConfig: {
+            mode: FunctionCallingConfigMode.AUTO // 允许模型自动决定是否调用工具
+          }
+        }
+      }
+    }
+
+    if (safetySettings) {
+      generateContentConfig.safetySettings = safetySettings
+    }
+
+    // 3. 一次性创建完整的 requestParams
+    const requestParams: GenerateContentParameters = {
+      model: modelId,
+      contents: formattedParts.contents,
+      config: generateContentConfig
+    }
+
+    // 发送流式请求
+    const result = await this.genAI.models.generateContentStream({
+      ...requestParams,
+      model: this.ensureVertexModelName(requestParams.model as string)
+    })
+
+    // 状态变量
+    let buffer = ''
+    let isInThinkTag = false
+    let toolUseDetected = false
+    let usageMetadata: GenerateContentResponseUsageMetadata | undefined
+    let isNewThoughtFormatDetected = modelConfig.reasoning === true
+
+    // 流处理循环
+    for await (const chunk of result) {
+      // 处理用量统计
+      if (chunk.usageMetadata) {
+        usageMetadata = chunk.usageMetadata
+      }
+
+      // 检查是否包含函数调用
+      if (chunk.candidates && chunk.candidates[0]?.content?.parts?.[0]?.functionCall) {
+        const functionCall = chunk.candidates[0].content.parts[0].functionCall
+        const functionName = functionCall.name
+        const functionArgs = functionCall.args || {}
+        const toolCallId = `gemini-tool-${Date.now()}`
+
+        toolUseDetected = true
+
+        // 发送工具调用开始事件
+        yield createStreamEvent.toolCallStart(toolCallId, functionName || '')
+
+        // 发送工具调用参数
+        const argsString = JSON.stringify(functionArgs)
+        yield createStreamEvent.toolCallChunk(toolCallId, argsString)
+
+        // 发送工具调用结束事件
+        yield createStreamEvent.toolCallEnd(toolCallId, argsString)
+
+        // 设置停止原因为工具使用
+        break
+      }
+
+      // 处理内容块
+      let content = ''
+      let thoughtContent = ''
+
+      // 处理文本和图像内容
+      if (chunk.candidates && chunk.candidates[0]?.content?.parts) {
+        for (const part of chunk.candidates[0].content.parts) {
+          // 检查是否是思考内容 (新格式)
+          if ((part as any).thought === true && part.text) {
+            isNewThoughtFormatDetected = true
+            thoughtContent += part.text
+          } else if (part.text) {
+            content += part.text
+          } else if (part.inlineData && part.inlineData.data && part.inlineData.mimeType) {
+            // 处理图像数据
+            yield createStreamEvent.imageData({
+              data: part.inlineData.data,
+              mimeType: part.inlineData.mimeType
+            })
+          }
+        }
+      } else {
+        // 兼容处理
+        content = chunk.text || ''
+      }
+
+      // 如果检测到思考内容，直接发送
+      if (thoughtContent) {
+        yield createStreamEvent.reasoning(thoughtContent)
+      }
+
+      if (!content) continue
+
+      if (isNewThoughtFormatDetected) {
+        yield createStreamEvent.text(content)
+      } else {
+        buffer += content
+
+        if (buffer.includes('<think>') && !isInThinkTag) {
+          const thinkStart = buffer.indexOf('<think>')
+          if (thinkStart > 0) {
+            yield createStreamEvent.text(buffer.substring(0, thinkStart))
+          }
+          buffer = buffer.substring(thinkStart + 7)
+          isInThinkTag = true
+        }
+
+        if (isInThinkTag && buffer.includes('</think>')) {
+          const thinkEnd = buffer.indexOf('</think>')
+          const reasoningContent = buffer.substring(0, thinkEnd)
+          if (reasoningContent) {
+            yield createStreamEvent.reasoning(reasoningContent)
+          }
+          buffer = buffer.substring(thinkEnd + 8)
+          isInThinkTag = false
+        }
+
+        if (!isInThinkTag && buffer) {
+          yield createStreamEvent.text(buffer)
+          buffer = ''
+        }
+      }
+    }
+
+    if (usageMetadata) {
+      yield createStreamEvent.usage({
+        prompt_tokens: usageMetadata.promptTokenCount || 0,
+        completion_tokens: usageMetadata.candidatesTokenCount || 0,
+        total_tokens: usageMetadata.totalTokenCount || 0
+      })
+    }
+
+    // 处理剩余缓冲区内容
+    if (!isNewThoughtFormatDetected && buffer) {
+      if (isInThinkTag) {
+        yield createStreamEvent.reasoning(buffer)
+      } else {
+        yield createStreamEvent.text(buffer)
+      }
+    }
+
+    // 发送停止事件
+    yield createStreamEvent.stop(toolUseDetected ? 'tool_use' : 'complete')
+  }
+
+  /**
+   * 处理图片生成模型的流式输出
+   */
+  private async *handleImageGenerationStream(
+    messages: ChatMessage[],
+    modelId: string,
+    temperature?: number,
+    maxTokens?: number
+  ): AsyncGenerator<LLMCoreStreamEvent> {
+    try {
+      // 提取用户消息并构建parts数组
+      const userMessage = messages.findLast((msg) => msg.role === 'user')
+      if (!userMessage) {
+        throw new Error('No user message found for image generation')
+      }
+
+      // 构建包含文本和图片的parts数组，参考formatVertexMessages的逻辑
+      const parts: Part[] = []
+
+      if (typeof userMessage.content === 'string') {
+        // 处理纯文本消息
+        if (userMessage.content.trim() !== '') {
+          parts.push({ text: userMessage.content })
+        }
+      } else if (Array.isArray(userMessage.content)) {
+        // 处理多模态消息（带图片等）
+        for (const part of userMessage.content) {
+          if (part.type === 'text') {
+            // 只添加非空文本
+            if (part.text && part.text.trim() !== '') {
+              parts.push({ text: part.text })
+            }
+          } else if (part.type === 'image_url' && part.image_url) {
+            // 处理图片（假设是 base64 格式）
+            const matches = part.image_url.url.match(/^data:([^;]+);base64,(.+)$/)
+            if (matches && matches.length === 3) {
+              const mimeType = matches[1]
+              const base64Data = matches[2]
+              parts.push({
+                inlineData: {
+                  data: base64Data,
+                  mimeType: mimeType
+                }
+              })
+            }
+          }
+        }
+      }
+
+      // 如果没有有效的parts，抛出错误
+      if (parts.length === 0) {
+        throw new Error('No valid content found for image generation')
+      }
+
+      // 发送生成请求
+      const result = await this.genAI.models.generateContentStream({
+        model: this.ensureVertexModelName(modelId),
+        contents: [{ role: 'user', parts }],
+        config: this.getGenerateContentConfig(temperature, maxTokens, modelId, false) // 图像生成不需要reasoning
+      })
+
+      // 处理流式响应
+      for await (const chunk of result) {
+        if (chunk.candidates && chunk.candidates[0]?.content?.parts) {
+          for (const part of chunk.candidates[0].content.parts) {
+            if (part.text) {
+              // 输出文本内容
+              yield createStreamEvent.text(part.text)
+            } else if (part.inlineData) {
+              // 输出图像数据
+              yield createStreamEvent.imageData({
+                data: part.inlineData.data || '',
+                mimeType: part.inlineData.mimeType || ''
+              })
+            }
+          }
+        }
+      }
+
+      // 发送停止事件
+      yield createStreamEvent.stop('complete')
+    } catch (error) {
+      console.error('Image generation stream error:', error)
+      yield createStreamEvent.error(
+        error instanceof Error ? error.message : 'Image generation failed'
+      )
+      yield createStreamEvent.stop('error')
+    }
+  }
+
+  async getEmbeddings(modelId: string, texts: string[]): Promise<number[][]> {
+    if (!this.genAI) throw new Error('Google Generative AI client is not initialized')
+    // Vertex embedContent 支持批量输入
+    const resp = await this.genAI.models.embedContent({
+      model: this.ensureVertexModelName(modelId),
+      contents: texts.map((text) => ({
+        parts: [{ text }]
+      }))
+    })
+    // resp.embeddings?: ContentEmbedding[]
+    if (resp && Array.isArray(resp.embeddings)) {
+      return resp.embeddings.map((e) => (Array.isArray(e.values) ? e.values : []))
+    }
+    // 若无返回，抛出异常
+    throw new Error('Vertex AI embedding API did not return embeddings')
+  }
+}
diff --git a/src/renderer/settings/components/ModelProviderSettingsDetail.vue b/src/renderer/settings/components/ModelProviderSettingsDetail.vue
index e6f1d4545..cf8038f2d 100644
--- a/src/renderer/settings/components/ModelProviderSettingsDetail.vue
+++ b/src/renderer/settings/components/ModelProviderSettingsDetail.vue
@@ -16,6 +16,15 @@
 
         <Separator />
 
+        <VertexProviderSettingsDetail
+          v-if="provider.apiType === 'vertex'"
+          :provider="provider as VERTEX_PROVIDER"
+          @config-updated="handleConfigChanged"
+          @validate-provider="validateApiKey"
+        />
+
+        <Separator v-if="provider.apiType === 'vertex'" />
+
         <!-- Azure特殊配置 -->
         <AzureProviderConfig
           v-if="provider.id === 'azure-openai'"
@@ -81,11 +90,12 @@
 import { computed, ref, watch, reactive } from 'vue'
 import { useProviderStore } from '@/stores/providerStore'
 import { useModelStore } from '@/stores/modelStore'
-import type { LLM_PROVIDER, RENDERER_MODEL_META } from '@shared/presenter'
+import type { LLM_PROVIDER, RENDERER_MODEL_META, VERTEX_PROVIDER } from '@shared/presenter'
 import { ScrollArea } from '@shadcn/components/ui/scroll-area'
 import ProviderApiConfig from './ProviderApiConfig.vue'
 import AzureProviderConfig from './AzureProviderConfig.vue'
 import GeminiSafetyConfig from './GeminiSafetyConfig.vue'
+import VertexProviderSettingsDetail from './VertexProviderSettingsDetail.vue'
 import ProviderRateLimitConfig from './ProviderRateLimitConfig.vue'
 import ModelScopeMcpSync from './ModelScopeMcpSync.vue'
 import ProviderModelManager from './ProviderModelManager.vue'
diff --git a/src/renderer/settings/components/VertexProviderSettingsDetail.vue b/src/renderer/settings/components/VertexProviderSettingsDetail.vue
new file mode 100644
index 000000000..1d44a6bde
--- /dev/null
+++ b/src/renderer/settings/components/VertexProviderSettingsDetail.vue
@@ -0,0 +1,216 @@
+<template>
+  <div class="flex flex-col gap-4">
+    <div class="flex flex-col items-start gap-2">
+      <Label :for="`${provider.id}-projectId`" class="flex-1 cursor-pointer">
+        {{ t('settings.provider.vertexProjectId') }}
+      </Label>
+      <Input
+        :id="`${provider.id}-projectId`"
+        :model-value="projectId"
+        :placeholder="t('settings.provider.vertexProjectIdPlaceholder')"
+        @blur="handleProjectIdChange(String($event.target?.value || ''))"
+        @keyup.enter="handleProjectIdChange(projectId)"
+        @update:model-value="projectId = String($event)"
+      />
+    </div>
+
+    <div class="flex flex-col items-start gap-2">
+      <Label :for="`${provider.id}-location`" class="flex-1 cursor-pointer">
+        {{ t('settings.provider.vertexLocation') }}
+      </Label>
+      <Input
+        :id="`${provider.id}-location`"
+        :model-value="location"
+        :placeholder="t('settings.provider.vertexLocationPlaceholder')"
+        @blur="handleLocationChange(String($event.target?.value || ''))"
+        @keyup.enter="handleLocationChange(location)"
+        @update:model-value="location = String($event)"
+      />
+    </div>
+
+    <div class="grid grid-cols-1 md:grid-cols-2 gap-4">
+      <div class="flex flex-col items-start gap-2">
+        <Label :for="`${provider.id}-apiVersion`" class="flex-1 cursor-pointer">
+          {{ t('settings.provider.vertexApiVersion') }}
+        </Label>
+        <Select v-model="apiVersion" @update:model-value="handleApiVersionChange">
+          <SelectTrigger class="w-full">
+            <SelectValue placeholder="v1" />
+          </SelectTrigger>
+          <SelectContent>
+            <SelectItem value="v1">v1</SelectItem>
+            <SelectItem value="v1beta1">v1beta1</SelectItem>
+          </SelectContent>
+        </Select>
+      </div>
+
+      <div class="flex flex-col items-start gap-2">
+        <Label :for="`${provider.id}-endpointMode`" class="flex-1 cursor-pointer">
+          {{ t('settings.provider.vertexEndpointMode') }}
+        </Label>
+        <Select v-model="endpointMode" @update:model-value="handleEndpointModeChange">
+          <SelectTrigger class="w-full">
+            <SelectValue :placeholder="t('settings.provider.vertexEndpointMode')" />
+          </SelectTrigger>
+          <SelectContent>
+            <SelectItem value="standard">
+              {{ t('settings.provider.vertexEndpointStandard') }}
+            </SelectItem>
+            <SelectItem value="express">
+              {{ t('settings.provider.vertexEndpointExpress') }}
+            </SelectItem>
+          </SelectContent>
+        </Select>
+      </div>
+    </div>
+
+    <div class="flex flex-col items-start gap-2">
+      <Label :for="`${provider.id}-serviceEmail`" class="flex-1 cursor-pointer">
+        {{ t('settings.provider.vertexServiceEmail') }}
+      </Label>
+      <Input
+        :id="`${provider.id}-serviceEmail`"
+        :model-value="accountClientEmail"
+        :placeholder="t('settings.provider.vertexServiceEmailPlaceholder')"
+        @blur="handleServiceEmailChange(String($event.target?.value || ''))"
+        @keyup.enter="handleServiceEmailChange(accountClientEmail)"
+        @update:model-value="accountClientEmail = String($event)"
+      />
+    </div>
+
+    <div class="flex flex-col items-start gap-2">
+      <Label :for="`${provider.id}-privateKey`" class="flex-1 cursor-pointer">
+        {{ t('settings.provider.vertexPrivateKey') }}
+      </Label>
+      <div class="relative w-full">
+        <Input
+          :id="`${provider.id}-privateKey`"
+          :model-value="accountPrivateKey"
+          :type="showPrivateKey ? 'text' : 'password'"
+          :placeholder="t('settings.provider.vertexPrivateKeyPlaceholder')"
+          style="padding-right: 2.5rem !important"
+          @blur="handlePrivateKeyChange(String($event.target?.value || ''))"
+          @keyup.enter="handlePrivateKeyChange(accountPrivateKey)"
+          @update:model-value="accountPrivateKey = String($event)"
+        />
+        <Button
+          variant="ghost"
+          size="sm"
+          class="absolute right-2 top-1/2 transform -translate-y-1/2 h-7 w-7 p-0 hover:bg-transparent"
+          @click="showPrivateKey = !showPrivateKey"
+        >
+          <Icon
+            :icon="showPrivateKey ? 'lucide:eye-off' : 'lucide:eye'"
+            class="w-4 h-4 text-muted-foreground hover:text-foreground"
+          />
+        </Button>
+      </div>
+    </div>
+
+    <div class="flex flex-row gap-2">
+      <Button
+        variant="outline"
+        size="sm"
+        class="text-xs text-normal rounded-lg"
+        @click="emit('validate-provider')"
+      >
+        <Icon icon="lucide:check-check" class="w-4 h-4 text-muted-foreground" />{{
+          t('settings.provider.verifyKey')
+        }}
+      </Button>
+    </div>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, watch } from 'vue'
+import { useI18n } from 'vue-i18n'
+import { Label } from '@shadcn/components/ui/label'
+import { Input } from '@shadcn/components/ui/input'
+import {
+  Select,
+  SelectContent,
+  SelectItem,
+  SelectTrigger,
+  SelectValue
+} from '@shadcn/components/ui/select'
+import { Button } from '@shadcn/components/ui/button'
+import { Icon } from '@iconify/vue'
+import type { VERTEX_PROVIDER } from '@shared/presenter'
+import { useProviderStore } from '@/stores/providerStore'
+
+const props = defineProps<{
+  provider: VERTEX_PROVIDER
+}>()
+
+const emit = defineEmits<{
+  'config-updated': []
+  'validate-provider': []
+}>()
+
+const { t } = useI18n()
+const providerStore = useProviderStore()
+
+const projectId = ref(props.provider.projectId || '')
+const location = ref(props.provider.location || '')
+const apiVersion = ref(props.provider.apiVersion || 'v1')
+const endpointMode = ref(props.provider.endpointMode || 'standard')
+const accountClientEmail = ref(props.provider.accountClientEmail || '')
+const accountPrivateKey = ref(props.provider.accountPrivateKey || '')
+const showPrivateKey = ref(false)
+
+watch(
+  () => props.provider,
+  (next) => {
+    projectId.value = next.projectId || ''
+    location.value = next.location || ''
+    apiVersion.value = next.apiVersion || 'v1'
+    endpointMode.value = next.endpointMode || 'standard'
+    accountClientEmail.value = next.accountClientEmail || ''
+    accountPrivateKey.value = next.accountPrivateKey || ''
+  },
+  { deep: true }
+)
+
+const updateConfig = async (updates: Partial<VERTEX_PROVIDER>) => {
+  await providerStore.updateVertexProviderConfig(props.provider.id, updates)
+  emit('config-updated')
+}
+
+const handleProjectIdChange = async (value: string) => {
+  const nextValue = value.trim()
+  projectId.value = nextValue
+  await updateConfig({ projectId: nextValue })
+}
+
+const handleLocationChange = async (value: string) => {
+  const nextValue = value.trim()
+  location.value = nextValue
+  await updateConfig({ location: nextValue })
+}
+
+const handleApiVersionChange = async (value: any) => {
+  if (value && typeof value === 'string') {
+    apiVersion.value = value as 'v1' | 'v1beta1'
+    await updateConfig({ apiVersion: apiVersion.value })
+  }
+}
+
+const handleEndpointModeChange = async (value: any) => {
+  if (value && typeof value === 'string') {
+    endpointMode.value = value as 'standard' | 'express'
+    await updateConfig({ endpointMode: endpointMode.value })
+  }
+}
+
+const handleServiceEmailChange = async (value: string) => {
+  const nextValue = value.trim()
+  accountClientEmail.value = nextValue
+  await updateConfig({ accountClientEmail: nextValue })
+}
+
+const handlePrivateKeyChange = async (value: string) => {
+  accountPrivateKey.value = value
+  await updateConfig({ accountPrivateKey: value })
+}
+</script>
diff --git a/src/renderer/src/components/icons/ModelIcon.vue b/src/renderer/src/components/icons/ModelIcon.vue
index 32596890f..1f85ccd79 100644
--- a/src/renderer/src/components/icons/ModelIcon.vue
+++ b/src/renderer/src/components/icons/ModelIcon.vue
@@ -106,6 +106,7 @@ const icons = {
   upstage: upstageColorIcon,
   vercel: vercelColorIcon,
   vertexai: vertexaiColorIcon,
+  vertex: vertexaiColorIcon,
   vidu: viduColorIcon,
   viggle: viggleColorIcon,
   tiangong: tiangongColorIcon,
diff --git a/src/renderer/src/i18n/da-DK/settings.json b/src/renderer/src/i18n/da-DK/settings.json
index f8a42dcc3..ca030a06e 100644
--- a/src/renderer/src/i18n/da-DK/settings.json
+++ b/src/renderer/src/i18n/da-DK/settings.json
@@ -489,7 +489,19 @@
     "operationSuccess": "Handling gennemført",
     "settingsApplied": "Indstillinger anvendt",
     "bedrockLimitTip": "* Kun Anthropic Claude (Opus, Sonnet, Haiku) understøttes",
-    "bedrockVerifyTip": "DeepChat bruger Claude 3.5 Sonnet til verifikation. Hvis du ikke har adgang, mislykkes testen – dette påvirker ikke andre modeller."
+    "bedrockVerifyTip": "DeepChat bruger Claude 3.5 Sonnet til verifikation. Hvis du ikke har adgang, mislykkes testen – dette påvirker ikke andre modeller.",
+    "vertexApiVersion": "API version",
+    "vertexEndpointExpress": "Express (globalt slutpunkt)",
+    "vertexEndpointMode": "slutpunktstilstand",
+    "vertexEndpointStandard": "Standard (områdeslutpunkt)",
+    "vertexLocation": "areal",
+    "vertexLocationPlaceholder": "Indtast venligst en region (f.eks. us-central1)",
+    "vertexPrivateKey": "Tjenestekonto privat nøgle",
+    "vertexPrivateKeyPlaceholder": "Indsæt den private nøgle (understøtter enkelt linje \\n)",
+    "vertexProjectId": "Projekt ID",
+    "vertexProjectIdPlaceholder": "Indtast dit Google Cloud-projekt-id",
+    "vertexServiceEmailPlaceholder": "Indtast venligst tjenestekontoens e-mail",
+    "vertexServiceEmail": "Servicekonto-e-mail"
   },
   "knowledgeBase": {
     "title": "Indstillinger for vidensbase",
diff --git a/src/renderer/src/i18n/en-US/settings.json b/src/renderer/src/i18n/en-US/settings.json
index 1d06f9205..f7cb503be 100644
--- a/src/renderer/src/i18n/en-US/settings.json
+++ b/src/renderer/src/i18n/en-US/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "Please enter AWS Access Key ID",
     "secretAccessKeyPlaceholder": "Please enter AWS Secret Access Key",
     "regionPlaceholder": "Please enter AWS region",
+    "vertexProjectId": "Project ID",
+    "vertexProjectIdPlaceholder": "Please enter your Google Cloud project ID",
+    "vertexLocation": "area",
+    "vertexLocationPlaceholder": "Please enter a region (e.g. us-central1)",
+    "vertexServiceEmail": "Service account email",
+    "vertexServiceEmailPlaceholder": "Please enter the service account email",
+    "vertexPrivateKey": "Service account private key",
+    "vertexPrivateKeyPlaceholder": "Paste the private key (supports single line \\n)",
+    "vertexApiVersion": "API version",
+    "vertexEndpointMode": "endpoint mode",
+    "vertexEndpointStandard": "Standard (region endpoint)",
+    "vertexEndpointExpress": "Express (global endpoint)",
     "verifyKey": "Verify Key",
     "howToGet": "How to get",
     "getKeyTip": "Please visit",
diff --git a/src/renderer/src/i18n/fa-IR/settings.json b/src/renderer/src/i18n/fa-IR/settings.json
index 5afb30c5c..0fc4f9afd 100644
--- a/src/renderer/src/i18n/fa-IR/settings.json
+++ b/src/renderer/src/i18n/fa-IR/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "لطفاً AWS Access Key ID را وارد کنید",
     "secretAccessKeyPlaceholder": "لطفاً AWS Secret Access Key را وارد کنید",
     "regionPlaceholder": "لطفاً منطقه AWS را وارد کنید",
+    "vertexProjectId": "شناسه پروژه",
+    "vertexProjectIdPlaceholder": "لطفا شناسه پروژه Google Cloud خود را وارد کنید",
+    "vertexLocation": "منطقه",
+    "vertexLocationPlaceholder": "لطفاً یک منطقه (به عنوان مثال us-central1) را وارد کنید",
+    "vertexServiceEmail": "ایمیل حساب سرویس",
+    "vertexServiceEmailPlaceholder": "لطفا ایمیل حساب سرویس را وارد کنید",
+    "vertexPrivateKey": "کلید خصوصی حساب سرویس",
+    "vertexPrivateKeyPlaceholder": "کلید خصوصی را جایگذاری کنید (از یک خط پشتیبانی می کند \\n)",
+    "vertexApiVersion": "نسخه API",
+    "vertexEndpointMode": "حالت نقطه پایانی",
+    "vertexEndpointStandard": "استاندارد (نقطه پایانی منطقه)",
+    "vertexEndpointExpress": "اکسپرس (نقطه پایانی جهانی)",
     "verifyKey": "پذیرش کلید",
     "howToGet": "چگونه دریافت کنیم",
     "getKeyTip": "لطفاً به",
diff --git a/src/renderer/src/i18n/fr-FR/settings.json b/src/renderer/src/i18n/fr-FR/settings.json
index 09873a00e..ebfd842e1 100644
--- a/src/renderer/src/i18n/fr-FR/settings.json
+++ b/src/renderer/src/i18n/fr-FR/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "Veuillez entrer l'AWS Access Key ID",
     "secretAccessKeyPlaceholder": "Veuillez entrer l'AWS Secret Access Key",
     "regionPlaceholder": "Veuillez entrer la région AWS",
+    "vertexProjectId": "ID du projet",
+    "vertexProjectIdPlaceholder": "Veuillez saisir l'ID de votre projet Google Cloud",
+    "vertexLocation": "zone",
+    "vertexLocationPlaceholder": "Veuillez entrer une région (par exemple us-central1)",
+    "vertexServiceEmail": "E-mail du compte de service",
+    "vertexServiceEmailPlaceholder": "Veuillez saisir l'e-mail du compte de service",
+    "vertexPrivateKey": "Clé privée du compte de service",
+    "vertexPrivateKeyPlaceholder": "Collez la clé privée (prend en charge une seule ligne \\n)",
+    "vertexApiVersion": "Version API",
+    "vertexEndpointMode": "mode point de terminaison",
+    "vertexEndpointStandard": "Standard (point de terminaison de la région)",
+    "vertexEndpointExpress": "Express (point de terminaison mondial)",
     "verifyKey": "Vérifier la clé",
     "howToGet": "Comment obtenir",
     "getKeyTip": "Veuillez visiter",
diff --git a/src/renderer/src/i18n/ja-JP/settings.json b/src/renderer/src/i18n/ja-JP/settings.json
index ad179db46..4a9728c75 100644
--- a/src/renderer/src/i18n/ja-JP/settings.json
+++ b/src/renderer/src/i18n/ja-JP/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "AWS Access Key IDを入力してください",
     "secretAccessKeyPlaceholder": "AWS Secret Access Keyを入力してください",
     "regionPlaceholder": "AWSリージョンを入力してください",
+    "vertexProjectId": "プロジェクトID",
+    "vertexProjectIdPlaceholder": "Google Cloud プロジェクト ID を入力してください",
+    "vertexLocation": "エリア",
+    "vertexLocationPlaceholder": "地域を入力してください (例: us-central1)",
+    "vertexServiceEmail": "サービスアカウントのメールアドレス",
+    "vertexServiceEmailPlaceholder": "サービスアカウントのメールアドレスを入力してください",
+    "vertexPrivateKey": "サービスアカウントの秘密キー",
+    "vertexPrivateKeyPlaceholder": "秘密キーを貼り付けます (単一行をサポート\\n)",
+    "vertexApiVersion": "API バージョン",
+    "vertexEndpointMode": "エンドポイントモード",
+    "vertexEndpointStandard": "標準（リージョンエンドポイント）",
+    "vertexEndpointExpress": "Express（グローバルエンドポイント）",
     "verifyKey": "キーを検証",
     "howToGet": "取得方法",
     "getKeyTip": "以下へアクセスしてください",
diff --git a/src/renderer/src/i18n/ko-KR/settings.json b/src/renderer/src/i18n/ko-KR/settings.json
index 8d4546c05..cdcf336f2 100644
--- a/src/renderer/src/i18n/ko-KR/settings.json
+++ b/src/renderer/src/i18n/ko-KR/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "AWS Access Key ID를 입력하세요",
     "secretAccessKeyPlaceholder": "AWS Secret Access Key를 입력하세요",
     "regionPlaceholder": "AWS 리전을 입력하세요",
+    "vertexProjectId": "프로젝트 ID",
+    "vertexProjectIdPlaceholder": "Google Cloud 프로젝트 ID를 입력하세요.",
+    "vertexLocation": "영역",
+    "vertexLocationPlaceholder": "지역을 입력하세요(예: us-central1).",
+    "vertexServiceEmail": "서비스 계정 이메일",
+    "vertexServiceEmailPlaceholder": "서비스 계정 이메일을 입력하세요.",
+    "vertexPrivateKey": "서비스 계정 비공개 키",
+    "vertexPrivateKeyPlaceholder": "개인 키 붙여넣기(한 줄 지원\\n)",
+    "vertexApiVersion": "API 버전",
+    "vertexEndpointMode": "엔드포인트 모드",
+    "vertexEndpointStandard": "표준(지역 엔드포인트)",
+    "vertexEndpointExpress": "Express(글로벌 엔드포인트)",
     "verifyKey": "키 확인",
     "howToGet": "얻는 방법",
     "getKeyTip": "다음으로 이동하세요",
diff --git a/src/renderer/src/i18n/pt-BR/settings.json b/src/renderer/src/i18n/pt-BR/settings.json
index bf34eade9..862d63c4a 100644
--- a/src/renderer/src/i18n/pt-BR/settings.json
+++ b/src/renderer/src/i18n/pt-BR/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "Por favor, insira o ID da Chave de Acesso da AWS",
     "secretAccessKeyPlaceholder": "Por favor, insira a Chave de Acesso Secreta da AWS",
     "regionPlaceholder": "Por favor, insira a região da AWS",
+    "vertexProjectId": "ID do projeto",
+    "vertexProjectIdPlaceholder": "Insira o ID do projeto do Google Cloud",
+    "vertexLocation": "área",
+    "vertexLocationPlaceholder": "Insira uma região (por exemplo, us-central1)",
+    "vertexServiceEmail": "E-mail da conta de serviço",
+    "vertexServiceEmailPlaceholder": "Insira o e-mail da conta de serviço",
+    "vertexPrivateKey": "Chave privada da conta de serviço",
+    "vertexPrivateKeyPlaceholder": "Cole a chave privada (suporta linha única \\n)",
+    "vertexApiVersion": "Versão da API",
+    "vertexEndpointMode": "modo de ponto final",
+    "vertexEndpointStandard": "Padrão (ponto final da região)",
+    "vertexEndpointExpress": "Expresso (endpoint global)",
     "verifyKey": "Verificar Chave",
     "howToGet": "Como obter",
     "getKeyTip": "Por favor, visite",
diff --git a/src/renderer/src/i18n/ru-RU/settings.json b/src/renderer/src/i18n/ru-RU/settings.json
index f24e12464..26f7dc715 100644
--- a/src/renderer/src/i18n/ru-RU/settings.json
+++ b/src/renderer/src/i18n/ru-RU/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "Введите AWS Access Key ID",
     "secretAccessKeyPlaceholder": "Введите AWS Secret Access Key",
     "regionPlaceholder": "Введите регион AWS",
+    "vertexProjectId": "Идентификатор проекта",
+    "vertexProjectIdPlaceholder": "Введите идентификатор проекта Google Cloud.",
+    "vertexLocation": "область",
+    "vertexLocationPlaceholder": "Пожалуйста, введите регион (например, us-central1)",
+    "vertexServiceEmail": "Адрес электронной почты сервисного аккаунта",
+    "vertexServiceEmailPlaceholder": "Пожалуйста, введите адрес электронной почты сервисного аккаунта",
+    "vertexPrivateKey": "Закрытый ключ сервисного аккаунта",
+    "vertexPrivateKeyPlaceholder": "Вставьте закрытый ключ (поддерживается одна строка \\n)",
+    "vertexApiVersion": "версия API",
+    "vertexEndpointMode": "режим конечной точки",
+    "vertexEndpointStandard": "Стандартный (конечная точка региона)",
+    "vertexEndpointExpress": "Экспресс (глобальная конечная точка)",
     "verifyKey": "Проверить ключ",
     "howToGet": "Как получить",
     "getKeyTip": "Перейдите по следующему адресу",
diff --git a/src/renderer/src/i18n/zh-CN/settings.json b/src/renderer/src/i18n/zh-CN/settings.json
index 357b3e6d7..3b299947f 100644
--- a/src/renderer/src/i18n/zh-CN/settings.json
+++ b/src/renderer/src/i18n/zh-CN/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "请输入AWS Access Key ID",
     "secretAccessKeyPlaceholder": "请输入AWS Secret Access Key",
     "regionPlaceholder": "请输入AWS区域",
+    "vertexProjectId": "项目 ID",
+    "vertexProjectIdPlaceholder": "请输入 Google Cloud 项目 ID",
+    "vertexLocation": "区域",
+    "vertexLocationPlaceholder": "请输入区域（例如 us-central1）",
+    "vertexServiceEmail": "服务账号邮箱",
+    "vertexServiceEmailPlaceholder": "请输入服务账号邮箱",
+    "vertexPrivateKey": "服务账号私钥",
+    "vertexPrivateKeyPlaceholder": "粘贴私钥（支持单行 \\n）",
+    "vertexApiVersion": "API 版本",
+    "vertexEndpointMode": "端点模式",
+    "vertexEndpointStandard": "标准（区域端点）",
+    "vertexEndpointExpress": "Express（全局端点）",
     "verifyKey": "验证密钥",
     "howToGet": "如何获取",
     "refreshingModels": "刷新中...",
diff --git a/src/renderer/src/i18n/zh-HK/settings.json b/src/renderer/src/i18n/zh-HK/settings.json
index ba3ec949a..c5a0b83ea 100644
--- a/src/renderer/src/i18n/zh-HK/settings.json
+++ b/src/renderer/src/i18n/zh-HK/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "請輸入AWS Access Key ID",
     "secretAccessKeyPlaceholder": "請輸入AWS Secret Access Key",
     "regionPlaceholder": "請輸入AWS區域",
+    "vertexProjectId": "項目 ID",
+    "vertexProjectIdPlaceholder": "請輸入 Google Cloud 項目 ID",
+    "vertexLocation": "區域",
+    "vertexLocationPlaceholder": "請輸入區域（例如 us-central1）",
+    "vertexServiceEmail": "服務賬號郵箱",
+    "vertexServiceEmailPlaceholder": "請輸入服務賬號郵箱",
+    "vertexPrivateKey": "服務賬號私鑰",
+    "vertexPrivateKeyPlaceholder": "粘貼私鑰（支持單行 \\n）",
+    "vertexApiVersion": "API 版本",
+    "vertexEndpointMode": "端點模式",
+    "vertexEndpointStandard": "標準（區域端點）",
+    "vertexEndpointExpress": "Express（全局端點）",
     "verifyKey": "驗證密鑰",
     "howToGet": "如何獲取",
     "getKeyTip": "請前往",
diff --git a/src/renderer/src/i18n/zh-TW/settings.json b/src/renderer/src/i18n/zh-TW/settings.json
index 5bc30e1ab..f887800e6 100644
--- a/src/renderer/src/i18n/zh-TW/settings.json
+++ b/src/renderer/src/i18n/zh-TW/settings.json
@@ -274,6 +274,18 @@
     "accessKeyIdPlaceholder": "請輸入 AWS Access Key ID",
     "secretAccessKeyPlaceholder": "請輸入 AWS Secret Access Key",
     "regionPlaceholder": "請輸入 AWS 區域",
+    "vertexProjectId": "項目 ID",
+    "vertexProjectIdPlaceholder": "請輸入 Google Cloud 項目 ID",
+    "vertexLocation": "區域",
+    "vertexLocationPlaceholder": "請輸入區域（例如 us-central1）",
+    "vertexServiceEmail": "服務賬號郵箱",
+    "vertexServiceEmailPlaceholder": "請輸入服務賬號郵箱",
+    "vertexPrivateKey": "服務賬號私鑰",
+    "vertexPrivateKeyPlaceholder": "粘貼私鑰（支持單行 \\n）",
+    "vertexApiVersion": "API 版本",
+    "vertexEndpointMode": "端點模式",
+    "vertexEndpointStandard": "標準（區域端點）",
+    "vertexEndpointExpress": "Express（全局端點）",
     "verifyKey": "驗證金鑰",
     "howToGet": "如何取得",
     "getKeyTip": "請前往",
diff --git a/src/renderer/src/stores/providerStore.ts b/src/renderer/src/stores/providerStore.ts
index a5f92bf98..428fb6067 100644
--- a/src/renderer/src/stores/providerStore.ts
+++ b/src/renderer/src/stores/providerStore.ts
@@ -3,7 +3,7 @@ import { defineStore } from 'pinia'
 import { usePresenter } from '@/composables/usePresenter'
 import { useIpcQuery } from '@/composables/useIpcQuery'
 import { CONFIG_EVENTS, PROVIDER_DB_EVENTS } from '@/events'
-import type { AWS_BEDROCK_PROVIDER, LLM_PROVIDER } from '@shared/presenter'
+import type { AWS_BEDROCK_PROVIDER, LLM_PROVIDER, VERTEX_PROVIDER } from '@shared/presenter'
 
 const PROVIDER_ORDER_KEY = 'providerOrder'
 const PROVIDER_TIMESTAMP_KEY = 'providerTimestamps'
@@ -293,6 +293,13 @@ export const useProviderStore = defineStore('provider', () => {
     return updateProviderConfig(providerId, updates)
   }
 
+  const updateVertexProviderConfig = async (
+    providerId: string,
+    updates: Partial<VERTEX_PROVIDER>
+  ) => {
+    return updateProviderConfig(providerId, updates)
+  }
+
   const checkProvider = async (providerId: string, modelId?: string) => {
     return llmP.check(providerId, modelId)
   }
@@ -400,6 +407,7 @@ export const useProviderStore = defineStore('provider', () => {
     addCustomProvider,
     removeProvider,
     updateAwsBedrockProviderConfig,
+    updateVertexProviderConfig,
     checkProvider,
     setAzureApiVersion,
     getAzureApiVersion,
diff --git a/src/shared/provider-operations.ts b/src/shared/provider-operations.ts
index dbdc08193..0187ddb28 100644
--- a/src/shared/provider-operations.ts
+++ b/src/shared/provider-operations.ts
@@ -46,7 +46,13 @@ export const REBUILD_REQUIRED_FIELDS = [
   'secretAccessKey', // AWS Bedrock
   'region', // AWS Bedrock
   'azureResourceName', // Azure
-  'azureApiVersion' // Azure
+  'azureApiVersion', // Azure
+  'projectId', // Vertex AI
+  'location', // Vertex AI
+  'accountPrivateKey', // Vertex AI
+  'accountClientEmail', // Vertex AI
+  'apiVersion', // Vertex AI
+  'endpointMode' // Vertex AI
 ] as const
 
 /**
diff --git a/src/shared/types/presenters/legacy.presenters.d.ts b/src/shared/types/presenters/legacy.presenters.d.ts
index 0191c7316..046599751 100644
--- a/src/shared/types/presenters/legacy.presenters.d.ts
+++ b/src/shared/types/presenters/legacy.presenters.d.ts
@@ -767,6 +767,15 @@ export type AWS_BEDROCK_PROVIDER = LLM_PROVIDER & {
   credential?: AwsBedrockCredential
 }
 
+export type VERTEX_PROVIDER = LLM_PROVIDER & {
+  projectId?: string
+  location?: string
+  accountPrivateKey?: string
+  accountClientEmail?: string
+  apiVersion?: 'v1' | 'v1beta1'
+  endpointMode?: 'standard' | 'express'
+}
+
 export interface AwsBedrockCredential {
   accessKeyId: string
   secretAccessKey: string
diff --git a/src/shared/types/presenters/llmprovider.presenter.d.ts b/src/shared/types/presenters/llmprovider.presenter.d.ts
index 476653e67..12eb25d89 100644
--- a/src/shared/types/presenters/llmprovider.presenter.d.ts
+++ b/src/shared/types/presenters/llmprovider.presenter.d.ts
@@ -91,6 +91,15 @@ export type AWS_BEDROCK_PROVIDER = LLM_PROVIDER & {
   credential?: AwsBedrockCredential
 }
 
+export type VERTEX_PROVIDER = LLM_PROVIDER & {
+  projectId?: string
+  location?: string
+  accountPrivateKey?: string
+  accountClientEmail?: string
+  apiVersion?: 'v1' | 'v1beta1'
+  endpointMode?: 'standard' | 'express'
+}
+
 export interface OllamaModel {
   name: string
   size: number