mirror of
https://github.com/wavetermdev/waveterm-docs-old.git
synced 2026-08-05 13:42:53 -07:00
added a doc for chat command
This commit is contained in:
@@ -0,0 +1,27 @@
|
||||
---
|
||||
title: 'ChatGPT Support'
|
||||
description: 'Talk to ChatGPT in the terminal'
|
||||
icon: 'robot'
|
||||
---
|
||||
|
||||
You can use the `chat` command to query AI chat models inside Wave terminal.
|
||||
|
||||
Usage
|
||||
```
|
||||
/chat tell me how to use the sed command
|
||||
```
|
||||
|
||||
Out of the box, this command will send a request through our cloud servers.
|
||||
If you'd prefer to query the OpenAI API directly, you can specify your own OpenAI key, as well as the model
|
||||
that you would like to use in the settings.
|
||||
|
||||
We also have support for changing the API base url if you want to use Azure models or other 3rd party API's.
|
||||
You can set the base url with the client:set command, like so:
|
||||
|
||||
```
|
||||
/client:set openaibaseurl="<BASE URL>"
|
||||
```
|
||||
The API you specify must have the same endpoints and same protocol as the OpenAI API.
|
||||
|
||||
In order to prevent abuse, you must have telemetry enabled to use the cloud servers. There is currently a rate limit of 200 requests per day.
|
||||
|
||||
Reference in New Issue
Block a user