diff --git a/config/config.example.json b/config/config.example.json
index 69e8feeae..67d04f66f 100644
--- a/config/config.example.json
+++ b/config/config.example.json
@@ -547,6 +547,7 @@
     "monitor_usb": true
   },
   "voice": {
+    "model_name": "",
     "echo_transcription": false
   },
   "gateway": {
diff --git a/docs/channels/telegram/README.md b/docs/channels/telegram/README.md
index a3e057ba4..5b4d6c76a 100644
--- a/docs/channels/telegram/README.md
+++ b/docs/channels/telegram/README.md
@@ -2,7 +2,7 @@
 
 # Telegram
 
-The Telegram channel uses long polling via the Telegram Bot API for bot-based communication. It supports text messages, media attachments (photos, voice, audio, documents), voice transcription via Groq Whisper, and built-in command handling.
+The Telegram channel uses long polling via the Telegram Bot API for bot-based communication. It supports text messages, media attachments (photos, voice, audio, documents), voice transcription ([setup](../../providers.md#voice-transcription)), and built-in command handling.
 
 ## Configuration
 
diff --git a/docs/channels/telegram/README.zh.md b/docs/channels/telegram/README.zh.md
index f50c712ce..6a7533582 100644
--- a/docs/channels/telegram/README.zh.md
+++ b/docs/channels/telegram/README.zh.md
@@ -2,7 +2,7 @@
 
 # Telegram
 
-Telegram Channel 通过 Telegram 机器人 API 使用长轮询实现基于机器人的通信。它支持文本消息、媒体附件（照片、语音、音频、文档）、通过 Groq Whisper 进行语音转录以及内置命令处理器。
+Telegram Channel 通过 Telegram 机器人 API 使用长轮询实现基于机器人的通信。它支持文本消息、媒体附件（照片、语音、音频、文档）、语音转录（配置见[提供商与模型配置](../../zh/providers.md#语音转录)），以及内置命令处理器。
 
 ## 配置
 
diff --git a/docs/providers.md b/docs/providers.md
index dde1814fb..3a740d3b8 100644
--- a/docs/providers.md
+++ b/docs/providers.md
@@ -5,7 +5,7 @@
 ### Providers
 
 > [!NOTE]
-> Groq provides free voice transcription via Whisper. If configured, audio messages from any channel will be automatically transcribed at the agent level.
+> Voice transcription can use a configured multimodal model via `voice.model_name`. Groq Whisper remains available as a fallback when no voice model is configured.
 
 | Provider     | Purpose                                 | Get API Key                                                  |
 | ------------ | --------------------------------------- | ------------------------------------------------------------ |
@@ -101,6 +101,33 @@ This design also enables **multi-agent support** with flexible provider selectio
 }
 ```
 
+#### Voice Transcription
+
+You can configure a dedicated model for audio transcription with `voice.model_name`. This lets you reuse existing multimodal providers that support audio input instead of relying only on Groq.
+
+If `voice.model_name` is not configured, PicoClaw will continue to fall back to Groq transcription when a Groq API key is available.
+
+```json
+{
+  "model_list": [
+    {
+      "model_name": "voice-gemini",
+      "model": "gemini/gemini-2.5-flash",
+      "api_key": "your-gemini-key"
+    }
+  ],
+  "voice": {
+    "model_name": "voice-gemini",
+    "echo_transcription": false
+  },
+  "providers": {
+    "groq": {
+      "api_key": "gsk_xxx"
+    }
+  }
+}
+```
+
 #### Vendor-Specific Examples
 
 **OpenAI**
@@ -344,6 +371,10 @@ picoclaw agent -m "Hello"
       "api_key": "gsk_xxx"
     }
   },
+  "voice": {
+    "model_name": "voice-gemini",
+    "echo_transcription": false
+  },
   "channels": {
     "telegram": {
       "enabled": true,
diff --git a/docs/zh/providers.md b/docs/zh/providers.md
index 9092e7dfe..e7b323ebf 100644
--- a/docs/zh/providers.md
+++ b/docs/zh/providers.md
@@ -5,7 +5,7 @@
 ### 提供商 (Providers)
 
 > [!NOTE]
-> Groq 通过 Whisper 提供免费的语音转录。如果配置了 Groq，任意渠道的音频消息都将在 Agent 层面自动转录为文字。
+> 语音转录现在可以通过 `voice.model_name` 指定的多模态模型完成；如果未配置语音模型，Groq Whisper 仍可作为回退方案。
 
 | 提供商               | 用途                         | 获取 API Key                                                         |
 | -------------------- | ---------------------------- | -------------------------------------------------------------------- |
@@ -99,6 +99,33 @@
 }
 ```
 
+#### 语音转录
+
+你可以通过 `voice.model_name` 为语音转录指定一个专用模型。这样可以直接复用已经配置好的、支持音频输入的多模态 provider，而不必只依赖 Groq。
+
+如果没有配置 `voice.model_name`，且存在 Groq API Key，PicoClaw 会继续回退到 Groq 转录。
+
+```json
+{
+  "model_list": [
+    {
+      "model_name": "voice-gemini",
+      "model": "gemini/gemini-2.5-flash",
+      "api_key": "your-gemini-key"
+    }
+  ],
+  "voice": {
+    "model_name": "voice-gemini",
+    "echo_transcription": false
+  },
+  "providers": {
+    "groq": {
+      "api_key": "gsk_xxx"
+    }
+  }
+}
+```
+
 #### 各厂商配置示例
 
 **OpenAI**
@@ -342,6 +369,10 @@ picoclaw agent -m "你好"
       "api_key": "gsk_xxx"
     }
   },
+  "voice": {
+    "model_name": "voice-gemini",
+    "echo_transcription": false
+  },
   "channels": {
     "telegram": {
       "enabled": true,