chenfei-wu/TaskMatrix · error

OpenAI API error.

Error message

OpenAI API error.

What it means

Returned (not raised) by executingLLM.execute when OpenAIWrapper.run reports failure. The wrapper catches every exception from openai.ChatCompletion.create and returns ('', False), so this string masks the true OpenAI error (auth, rate limit, timeout, quota, invalid deployment).

Source

Thrown at LowCodeLLM/src/executingLLM.py:42

"""

class executingLLM:
    def __init__(self, temperature) -> None:
        self.prefix = EXECUTING_LLM_PREFIX
        self.suffix = EXECUTING_LLM_SUFFIX
        self.LLM = OpenAIWrapper(temperature)
        self.messages = [{"role": "system", "content": "You are a helpful assistant."},
                         {"role": "system", "content": self.prefix}]

    def execute(self, current_prompt, history):
        ''' provide LLM the dialogue history and the current prompt to get response '''
        messages = self.messages + history
        messages.append({'role': 'user', "content": current_prompt + self.suffix})
        response, status = self.LLM.run(messages)
        if status:
            return response
        else:
            return "OpenAI API error."

View on GitHub (pinned to 4b7664f8d3)

Solutions

1. Log the exception inside OpenAIWrapper._post_request_chat instead of swallowing it, then rerun to see the real OpenAI error.
2. Verify OPENAIKEY is set and valid; if USE_AZURE=true verify API_BASE, API_VERSION, MODEL (deployment name).
3. Fix malformed history (list of {'role','content'} dicts) that the API would reject.
4. Retry with backoff for transient rate-limit/network errors.

Example fix

# before (openAIWrapper.py)
except Exception as e:
    return "", False
# after
except Exception as e:
    print('OpenAI call failed:', repr(e))
    return "", False

import os
assert os.environ.get('OPENAIKEY'), 'OPENAIKEY not set'
if os.environ.get('USE_AZURE','').lower() == 'true':
    for v in ('API_BASE','API_VERSION','MODEL'): assert os.environ.get(v), v + ' not set'

def is_llm_error(resp) -> bool:
    return isinstance(resp, str) and resp.strip() == 'OpenAI API error.'

for attempt in range(3):
    out = llm.execute(prompt, history)
    if not is_llm_error(out): break
    time.sleep(2 ** attempt)

When it happens

Trigger: Calling llm.execute with any conditions that make openai.ChatCompletion.create throw: OPENAIKEY unset/invalid, USE_AZURE=true but API_BASE/API_VERSION/MODEL unset or pointing at a wrong deployment, rate limit/quota exceeded, or network unreachable. Also triggered by malformed history entries missing 'role'/'content'.

Common situations: Env vars not exported into the Flask process; Azure deployment name not matching MODEL; expired billing/quota; using the deprecated openai<1.0 API against a newer key/provider that requires it (or vice versa).

Related errors

• OpenAI API error.
• failed to get_workflow, msg:%s, request data:%s
• failed to extend_workflow, msg:%s, request data:%s
• failed to execute, msg:%s, request data:%s
• internal errors

AI-assisted analysis of chenfei-wu/TaskMatrix@4b7664f8d3 (2026-08-27). Data as JSON: /api/errors/cd14c206820d7f00. Report an issue: GitHub.