CLI way of launching default ChatInterface (`load_chat`)

[vadimkantorov]

Rather than having to code up two-LOC python script as advised here: https://www.gradio.app/guides/creating-a-chatbot-fast#note-for-open-ai-api-compatible-endpoints

It would be nice to be able to fire up such most-basic chatbot interfaces using CLI with documented CLI arguments: python -m gradio openaichat --share --host ... --port ... or gradio openaichat --share --host ... --port .... The killer feature would be the --share option.

Currently it might be possible with a clutch:
python -c 'import sys, gradio as gr; gr.load_chat("http://{host}:{port}/v1/".format(host = sys.argv[-2], port = sys.argv[-1]), model=sys.argv[-3], token="***").launch(share=True)' "Qwen/Qwen3-8B" "127.0.0.1" "8080"

[abidlabs]

I like the idea of making gradio load-chat and gradio load cli commands, cc @aliabid94

[vadimkantorov]

It would also be important to be able to force text-only (non-multimodal) mode in CLI mode to work around explicitly problems like:

• Default gr.load_chat(...) with large and small inputs leads to model is not multimodal error in vllm serve #11331