Zhipu AI Translation Channel

What is it?

Zhipu AI (智谱 AI, also known as Zhipu) is a leading AI platform that provides the GLM series of models. Since version 3.74, pyVideoTrans has integrated Zhipu AI as a standalone translation channel. Additionally, Zhipu AI supports speech recognition (GLM-ASR) and dubbing (GLM-TTS).

Prerequisites

• pyVideoTrans v3.74 or higher (for translation)
• pyVideoTrans v3.93 or higher (for speech recognition and dubbing)
• A Zhipu AI platform account (to obtain an API key)

How to Get an API Key

Step 1: Sign up and log in to Zhipu AI

Go to https://www.bigmodel.cn and register with your phone number or email.

Step 2: Create an API Key

1. After logging in, visit https://www.bigmodel.cn/usercenter/apikeys
2. Click "Create API Key"
3. Give the key a name
4. Copy the generated API key and save it

Step 3: Claim free credits

Zhipu AI provides free credits for new users. Check your balance on the platform homepage.

Configuring Translation in pyVideoTrans

1. Open the menu bar → Translation Settings → Zhipu AI Settings
2. Enter the API key from Zhipu AI in the API KEY field
3. Select a model from the dropdown — glm-4-flash is recommended (free, fast), or manually enter other Zhipu AI model names
4. Click the Test button to verify the connection
5. Save after a successful test

Using Speech Recognition (GLM-ASR)

Starting from v3.93, a new speech recognition channel is available: Zhipu AI GLM-ASR.

1. Enter your API key in Translation Settings → Zhipu AI
2. Select the speech recognition model glm-asr-2512
3. On the main interface, select Zhipu AI GLM-ASR as the speech recognition channel
4. Supports speech recognition in 10+ languages

Using Dubbing (GLM-TTS)

Starting from v3.93, a new dubbing channel is available: Zhipu AI GLM-TTS.

1. Enter your API key in Translation Settings → Zhipu AI
2. On the main interface, select Zhipu AI GLM-TTS as the dubbing channel
3. Uses the glm-tts model for dubbing

Note: Default dubbing includes a "beep" watermark sound. To disable it, go to: Personal Center → Security Management → Watermark Management: https://www.bigmodel.cn/usercenter/safety-mgmt/watermark

Recommended Configuration

Setting	Recommended Value
API URL	https://open.bigmodel.cn/api/paas/v4/ (default, no changes needed)
API Key	Obtained from Zhipu AI platform
Translation Model	glm-4-flash (free, fast) or glm-4.5-flash
Max Output Tokens	4095 (default)

Troubleshooting

Issue	Cause	Solution
"Authentication Error"	API key incorrect	Log in to Zhipu AI and generate a new key
Dubbing has "beep" watermark	Watermark not disabled	Disable it at https://www.bigmodel.cn/usercenter/safety-mgmt/watermark
Speech recognition unavailable	Version too low	Upgrade to v3.93 or higher
Connection timeout	Network issue	Check your network connection
Rate limit errors	Too many requests	Increase the post-translation pause time