GLM4 Vision Integration
Recent:I have successfully load the vision understanding fuction of the one of the Chinese most powerful LLM GLM4 in COMFYUI. Anyuser could use their own API-KEY to use this fuction
GLM4_vision_node √
GLM4_node √
GLM3_trubo_node √
我在视频当中放的API是没法用的,大家要使用的话得自己获得自己的API
The API I included in the video cannot be used. If you want to use it, you need to obtain your own API.
GLM Vision is now integrated into this project. Due to the slow transmission rate of base64 in GLM4, we temporarily support uploading via URL. The function of this feature is to allow users to upload images for processing via a URL, then user could use this function to chat with GLM4 agent. This enhances the user experience and processing speed.
To use GLM4 Vision, follow the steps below:
Upload your image to a public URL. This can be done via different online hosting platforms. Ensure the image is accessible publicly. Use the provided function to upload the image URL. Upload your API key to the node. It’s very important to make sure that your API key is working properly.
GLM 4 is very powerful large language model capable of responding to and solving very complex problems
To use GLM4 , follow the steps below:
Upload your API key to the node. Then It’s very important to make sure that your API key is working properly. Then input your prompt. Use the queue to create the answer
GLM Vision is now integrated into this project. Due to the slow transmission rate of base64 in GLM4, we temporarily support uploading via URL. The function of this feature is to allow users to upload images for processing via a URL, then user could use this function to chat with GLM4 agent. This enhances the user experience and processing speed.
To use GLM4 , follow the steps below:
Upload your API key to the node. Then It’s very important to make sure that your API key is working properly. Then input your prompt. Use the queue to create the answer
pip install zhipuai
Please be careful when handling your API key to prevent misuse. We are still working to make the process more secure and efficient.
This method is in its beta phase. If you find any bug or have any suggestions, feel free to open an issue or a pull request. my email address is [email protected]