openclaw - 💡(How to fix) Fix [Feature]:Support openai-chat API type for providers [1 participants]

Severity: Blocks workflow — providers return "Unknown parameter: max_completion_tokens" error

Fix Plan

To add support for the openai-chat API type, follow these steps:

• Create a new API type in OpenClaw that maps to the /v1/chat/completions endpoint.
• Update the API request payload to conform to the chat completions format.
• Handle the response from the /v1/chat/completions endpoint.

Example code snippet in Python:

import requests

class OpenAIChatAPI:
    def __init__(self, api_url, api_key):
        self.api_url = api_url
        self.api_key = api_key

    def chat_completions(self, prompt, max_tokens):
        url = f"{self.api_url}/v1/chat/completions"
        headers = {
            "Authorization": f"Bearer {self.api_key}",
            "Content-Type": "application/json"
        }
        data = {
            "model": "your_model_name",
            "messages": [{"role": "user", "content": prompt}],
            "max_tokens": max_tokens
        }
        response = requests.post(url, headers=headers, json=data)
        return response.json()

# Example usage:
api = OpenAIChatAPI("https://api.example.com", "your_api_key")
response = api.chat_completions("Your prompt here", 1024)
print(response)

Verification

To verify that the fix worked, test the new API type with a Chinese AI provider that only supports the /v1/chat/completions endpoint. Check that the response from the API is correctly parsed and that the max_tokens parameter is no longer causing an error.

Extra Tips

• Make sure to update the API documentation to reflect the new API type and endpoint.
• Consider adding error handling for cases where the API provider returns an error or invalid response.