Handle VoidAI API Errors - VoidAI

Check the HTTP status and response body before reading a success payload. Gateway and upstream errors can use different fields; do not assume every failure includes an identical error envelope or a reference ID.

Status	What to check
400	Request shape, required parameters, model-specific constraints, and policy restrictions.
401	Missing, expired, or invalid credentials; see authentication.
403	Account, IP, plan, or model access restrictions.
404	Unknown resource/model or an unavailable endpoint.
429	Rate or usage limits; inspect any retry guidance.
5xx	Gateway/provider errors or temporary availability problems.

A public model listing does not prove that your account can call it. Claude access requires an additional grant. The moderation endpoint is unavailable to normal accounts, and video routes are not currently registered.

Handle SDK exceptions

import os
from openai import OpenAI, APIError, AuthenticationError, RateLimitError

client = OpenAI(
    api_key=os.environ['VOIDAI_API_KEY'],
    base_url='https://api.voidai.app/v1',
)
try:
    response = client.chat.completions.create(
        model='gpt-4o-mini',
        messages=[{'role': 'user', 'content': 'Hello'}],
    )
except AuthenticationError:
    print('Check the credential and authentication header.')
except RateLimitError:
    print('Inspect retry guidance and account limits before retrying.')
except APIError as error:
    print('Request failed with status:', error.status_code)

For transient failures, use bounded retries with exponential backoff and jitter. Avoid blindly retrying validation, authentication, or access failures. A failed connection does not always establish whether an inference request completed, so repeated requests can create additional usage.

Get support

Save the status, endpoint, time, and any reference_id or request/response ID returned. Remove API keys and sensitive prompt content before sharing diagnostics through support. Check account state in your dashboard.