How do I connect to a remote lmstudio server address?

[Cose-3D]

LM Studio is not reachable at ws://localhost:1234/llm (due to httpx.ConnectError: All connection attempts failed).
    Is LM Studio running?

By default the server is localhost and I can't find how to change it in the documentation.

[seamuss1]

Hello!
You will need to replace localhost with the IP address of your remote machine.
Localhost just refers to the computer that you are currently using.
You can find this by typing ipconfig in powershell or IP address in Linux, among other methods. Look for the IPv4 address.
Also you might need make an exception in the firewall to allow for traffic over the port (1234).
To test if traffic works you can run this PowerShell command
Test-NetConnection -ComputerName <target-computer> -Port 1234

Hope this helps!

[Cose-3D]

Hello! You will need to replace localhost with the IP address of your remote machine. Localhost just refers to the computer that you are currently using. You can find this by typing ipconfig in powershell or IP address in Linux, among other methods. Look for the IPv4 address. Also you might need make an exception in the firewall to allow for traffic over the port (1234). To test if traffic works you can run this PowerShell command Test-NetConnection -ComputerName <target-computer> -Port 1234

Hope this helps!

No offense but this is like a llm generated response.

model = lms.llm()

Where do I write the remote address

[seamuss1]

Hello,
Slight offense taken.

I was looking through the code for a solution, I found this: get_default_client()
Try this code out:

import lmstudio as lms

ip_address = '192.168.1.10'
port = 1234
lms.get_default_client(f"ws://{ip_address}:{port}/llm")
model = lms.llm()
model.complete("Once upon a time,")

[ncoghlan]

This info has now been added to the main SDK documentation: https://lmstudio.ai/docs/python/getting-started/project-setup#customizing-the-server-api-host-and-tcp-port

(The exact wording is still a work in progress)