diff --git a/README.md b/README.md
index bd9e5344b..a4ce80e8f 100644
--- a/README.md
+++ b/README.md
@@ -128,6 +128,8 @@ docker run -d --name telegram-search -p 3333:3333 ghcr.io/groupultra/telegram-se
 > [!IMPORTANT]
 > AI Embedding & LLM 设置现在在应用内**按账户**配置（设置 → API）。
 >
+> 支持多种 LLM 提供商，包括 [OpenAI](https://platform.openai.com/)、[MiniMax](https://www.minimaxi.com/) 等 OpenAI 兼容的 API 服务。在设置页面选择提供商后，API 地址和默认模型将自动填充。
+>
 > 请在修改完成 `.env` 文件后，再次执行 `docker compose -f docker-compose.yml up -d` 启动服务。
 
 以下环境变量全部为可选，如果不填写，则会使用默认值。
diff --git a/apps/web/src/locales/en.json b/apps/web/src/locales/en.json
index b421e4968..99796b086 100644
--- a/apps/web/src/locales/en.json
+++ b/apps/web/src/locales/en.json
@@ -194,6 +194,7 @@
     "llm": "Large Language Model",
     "llmModel": "LLM Model",
     "llmProvider": "LLM Provider",
+    "customProvider": "Custom",
     "visionLLM": "Vision LLM (Multimodal)",
     "visionLLMDescription": "Multimodal large language model for image understanding. Used to generate image descriptions for semantic search.",
     "login": "Login",
diff --git a/apps/web/src/locales/zh-CN.json b/apps/web/src/locales/zh-CN.json
index a3d08cecc..7f0308ccb 100644
--- a/apps/web/src/locales/zh-CN.json
+++ b/apps/web/src/locales/zh-CN.json
@@ -192,6 +192,7 @@
     "llm": "大语言模型",
     "llmModel": "LLM 模型",
     "llmProvider": "LLM 提供商",
+    "customProvider": "自定义",
     "visionLLM": "视觉大模型（多模态）",
     "visionLLMDescription": "多模态大语言模型，用于图片理解。用于生成图片描述以支持语义搜索。",
     "login": "登录",
diff --git a/apps/web/src/pages/settings.vue b/apps/web/src/pages/settings.vue
index 493b6caff..ab2273f7f 100644
--- a/apps/web/src/pages/settings.vue
+++ b/apps/web/src/pages/settings.vue
@@ -1,6 +1,8 @@
 <script setup lang="ts">
+import type { LLMProviderKey } from '@tg-search/core'
+
 import { useAccountStore, useBridge } from '@tg-search/client'
-import { CoreEventType } from '@tg-search/core'
+import { CoreEventType, LLM_PROVIDERS } from '@tg-search/core'
 import { storeToRefs } from 'pinia'
 import { computed, watch } from 'vue'
 import { useI18n } from 'vue-i18n'
@@ -24,6 +26,28 @@ const messageResolvers = [
 
 const embeddingDimensions = Object.values([1536, 1024, 768])
 
+const providerKeys = Object.keys(LLM_PROVIDERS) as LLMProviderKey[]
+
+const selectedProvider = computed({
+  get() {
+    const apiBase = accountSettings.value?.llm?.apiBase ?? ''
+    for (const key of providerKeys) {
+      if (apiBase === LLM_PROVIDERS[key].apiBase)
+        return key
+    }
+    return '' // custom
+  },
+  set(key: string) {
+    if (!key || !accountSettings.value)
+      return
+    const preset = LLM_PROVIDERS[key as LLMProviderKey]
+    if (!preset)
+      return
+    accountSettings.value.llm.apiBase = preset.apiBase
+    accountSettings.value.llm.model = preset.defaultModel
+  },
+})
+
 function buildDefaultMessageProcessing() {
   return {
     receiveMessages: { receiveAll: true, downloadMedia: true },
@@ -224,13 +248,28 @@ function updateConfig() {
               </div>
 
               <div class="grid gap-6">
+                <div class="space-y-2">
+                  <label class="text-sm font-medium">{{ t('settings.llmProvider') }}</label>
+                  <select
+                    v-model="selectedProvider"
+                    class="h-10 w-full flex border border-input rounded-md bg-background px-3 py-2 text-sm ring-offset-background disabled:cursor-not-allowed disabled:opacity-50 focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-offset-2 focus-visible:ring-ring"
+                  >
+                    <option v-for="key in providerKeys" :key="key" :value="key">
+                      {{ LLM_PROVIDERS[key].label }}
+                    </option>
+                    <option value="">
+                      {{ t('settings.customProvider') }}
+                    </option>
+                  </select>
+                </div>
+
                 <div class="grid gap-4 sm:grid-cols-2">
                   <div class="space-y-2">
                     <label class="text-sm font-medium">{{ t('settings.llmModel') }}</label>
                     <input
                       v-model="accountSettings.llm.model"
                       type="text"
-                      placeholder="gpt-4o-mini"
+                      :placeholder="LLM_PROVIDERS[selectedProvider as LLMProviderKey]?.defaultModel ?? 'gpt-4o-mini'"
                       class="h-10 w-full flex border border-input rounded-md bg-background px-3 py-2 text-sm ring-offset-background disabled:cursor-not-allowed file:border-0 file:bg-transparent file:text-sm placeholder:text-muted-foreground file:font-medium disabled:opacity-50 focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-offset-2 focus-visible:ring-ring"
                     >
                   </div>
@@ -239,7 +278,7 @@ function updateConfig() {
                     <input
                       v-model="accountSettings.llm.apiBase"
                       type="text"
-                      placeholder="https://api.openai.com/v1"
+                      :placeholder="LLM_PROVIDERS[selectedProvider as LLMProviderKey]?.apiBase ?? 'https://api.openai.com/v1'"
                       class="h-10 w-full flex border border-input rounded-md bg-background px-3 py-2 text-sm ring-offset-background disabled:cursor-not-allowed file:border-0 file:bg-transparent file:text-sm placeholder:text-muted-foreground file:font-medium disabled:opacity-50 focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-offset-2 focus-visible:ring-ring"
                     >
                   </div>
diff --git a/docs/README_EN.md b/docs/README_EN.md
index 10da1998a..9dd14c812 100644
--- a/docs/README_EN.md
+++ b/docs/README_EN.md
@@ -133,6 +133,8 @@ docker run -d --name telegram-search -p 3333:3333 ghcr.io/groupultra/telegram-se
 > [!IMPORTANT]
 > AI Embedding & LLM settings are now **per-account** in-app (Settings → API).
 >
+> Multiple LLM providers are supported, including [OpenAI](https://platform.openai.com/), [MiniMax](https://www.minimaxi.com/), and other OpenAI-compatible API services. Select a provider on the settings page to auto-fill the API endpoint and default model.
+>
 > Please restart the service after modifying the `.env` file by running `docker compose -f docker-compose.yml up -d`.
 
 All environment variables are optional. If not provided, the default values will be used.
diff --git a/packages/core/src/__test__/llm-providers.test.ts b/packages/core/src/__test__/llm-providers.test.ts
new file mode 100644
index 000000000..74b5176fb
--- /dev/null
+++ b/packages/core/src/__test__/llm-providers.test.ts
@@ -0,0 +1,84 @@
+import { describe, expect, it } from 'vitest'
+
+import { detectProviderFromApiBase, LLM_PROVIDER_KEYS, LLM_PROVIDERS } from '../llm-providers'
+
+describe('llm-providers', () => {
+  describe('lLM_PROVIDERS', () => {
+    it('should contain openai and minimax providers', () => {
+      expect(LLM_PROVIDERS).toHaveProperty('openai')
+      expect(LLM_PROVIDERS).toHaveProperty('minimax')
+    })
+
+    it('should have valid openai preset', () => {
+      const openai = LLM_PROVIDERS.openai
+      expect(openai.label).toBe('OpenAI')
+      expect(openai.apiBase).toBe('https://api.openai.com/v1')
+      expect(openai.defaultModel).toBe('gpt-4o-mini')
+      expect(openai.models.length).toBeGreaterThan(0)
+    })
+
+    it('should have valid minimax preset', () => {
+      const minimax = LLM_PROVIDERS.minimax
+      expect(minimax.label).toBe('MiniMax')
+      expect(minimax.apiBase).toBe('https://api.minimax.io/v1')
+      expect(minimax.defaultModel).toBe('MiniMax-M2.7')
+      expect(minimax.models).toContain('MiniMax-M2.7')
+      expect(minimax.models).toContain('MiniMax-M2.7-highspeed')
+      expect(minimax.models).toContain('MiniMax-M2.5')
+      expect(minimax.models).toContain('MiniMax-M2.5-highspeed')
+    })
+
+    it('should have unique apiBase for each provider', () => {
+      const bases = Object.values(LLM_PROVIDERS).map(p => p.apiBase)
+      expect(new Set(bases).size).toBe(bases.length)
+    })
+
+    it('should have non-empty label and defaultModel for each provider', () => {
+      for (const [key, preset] of Object.entries(LLM_PROVIDERS)) {
+        expect(preset.label, `${key} label`).toBeTruthy()
+        expect(preset.defaultModel, `${key} defaultModel`).toBeTruthy()
+        expect(preset.models.length, `${key} models`).toBeGreaterThan(0)
+      }
+    })
+
+    it('should include defaultModel in models list', () => {
+      for (const [key, preset] of Object.entries(LLM_PROVIDERS)) {
+        expect(preset.models, `${key} models should include defaultModel`).toContain(preset.defaultModel)
+      }
+    })
+  })
+
+  describe('lLM_PROVIDER_KEYS', () => {
+    it('should contain all provider keys', () => {
+      expect(LLM_PROVIDER_KEYS).toContain('openai')
+      expect(LLM_PROVIDER_KEYS).toContain('minimax')
+    })
+
+    it('should match Object.keys of LLM_PROVIDERS', () => {
+      expect(LLM_PROVIDER_KEYS).toEqual(Object.keys(LLM_PROVIDERS))
+    })
+  })
+
+  describe('detectProviderFromApiBase', () => {
+    it('should detect openai provider', () => {
+      expect(detectProviderFromApiBase('https://api.openai.com/v1')).toBe('openai')
+    })
+
+    it('should detect minimax provider', () => {
+      expect(detectProviderFromApiBase('https://api.minimax.io/v1')).toBe('minimax')
+    })
+
+    it('should return undefined for unknown URL', () => {
+      expect(detectProviderFromApiBase('https://api.example.com/v1')).toBeUndefined()
+    })
+
+    it('should return undefined for empty string', () => {
+      expect(detectProviderFromApiBase('')).toBeUndefined()
+    })
+
+    it('should not match partial URLs', () => {
+      expect(detectProviderFromApiBase('https://api.openai.com')).toBeUndefined()
+      expect(detectProviderFromApiBase('https://api.minimax.io')).toBeUndefined()
+    })
+  })
+})
diff --git a/packages/core/src/index.ts b/packages/core/src/index.ts
index e50a0ac8b..711e4bd60 100644
--- a/packages/core/src/index.ts
+++ b/packages/core/src/index.ts
@@ -4,6 +4,8 @@ export { initDrizzle } from './db'
 export type { CoreDB, InitDrizzleResult } from './db'
 export type * from './event-handlers'
 export { createCoreInstance, destroyCoreInstance } from './instance'
+export { detectProviderFromApiBase, LLM_PROVIDER_KEYS, LLM_PROVIDERS } from './llm-providers'
+export type { LLMProviderKey, LLMProviderPreset } from './llm-providers'
 export * from './models'
 export type * from './types'
 export { CoreEventType } from './types/events'
diff --git a/packages/core/src/llm-providers.ts b/packages/core/src/llm-providers.ts
new file mode 100644
index 000000000..45c5125f1
--- /dev/null
+++ b/packages/core/src/llm-providers.ts
@@ -0,0 +1,38 @@
+export interface LLMProviderPreset {
+  label: string
+  apiBase: string
+  defaultModel: string
+  models: string[]
+}
+
+export const LLM_PROVIDERS = {
+  openai: {
+    label: 'OpenAI',
+    apiBase: 'https://api.openai.com/v1',
+    defaultModel: 'gpt-4o-mini',
+    models: ['gpt-4o', 'gpt-4o-mini', 'gpt-4.1', 'gpt-4.1-mini', 'gpt-4.1-nano', 'o3-mini'],
+  },
+  minimax: {
+    label: 'MiniMax',
+    apiBase: 'https://api.minimax.io/v1',
+    defaultModel: 'MiniMax-M2.7',
+    models: ['MiniMax-M2.7', 'MiniMax-M2.7-highspeed', 'MiniMax-M2.5', 'MiniMax-M2.5-highspeed'],
+  },
+} as const satisfies Record<string, LLMProviderPreset>
+
+export type LLMProviderKey = keyof typeof LLM_PROVIDERS
+
+export const LLM_PROVIDER_KEYS = Object.keys(LLM_PROVIDERS) as LLMProviderKey[]
+
+/**
+ * Detect provider key from an API base URL.
+ * Returns `undefined` for unrecognised or custom endpoints.
+ */
+export function detectProviderFromApiBase(apiBase: string): LLMProviderKey | undefined {
+  for (const [key, preset] of Object.entries(LLM_PROVIDERS)) {
+    if (apiBase === preset.apiBase) {
+      return key as LLMProviderKey
+    }
+  }
+  return undefined
+}