Resolving API Key, Quota & Rate Limit Errors

Eviory OS operates on a BYOK (Bring Your Own Key) architecture. Because your browser communicates directly with the AI provider’s endpoint, errors usually indicate an issue on the provider account’s credit balance, key permissions, or rate limits.

Common Error Codes & Solutions

1. 401 Unauthorized / Incorrect API Key provided

• Cause: The API key entered in Settings contains typos, extra whitespace, or has been revoked on your provider dashboard.
• Fix:
  1. Go to your provider console (e.g., OpenRouter Keys or DeepSeek Console).
  2. Generate a fresh API key and copy it carefully.
  3. In Eviory OS, go to Settings → AI Integrations, replace the existing key, and hit Save.

2. 429 Too Many Requests / Quota Exceeded

• Cause:
  • Your account has run out of prepaid credits.
  • You hit the Tier 1 request-per-minute (RPM) limit on your account.
• Fix:
  1. Check your billing dashboard on the provider’s website. Ensure you have positive prepaid credit balance (often distinct from standard ChatGPT Plus or Claude Pro consumer subscriptions).
  2. If generating long batch campaigns, pause for 10–20 seconds before submitting the next prompt, or hot-swap to an alternative provider like DeepSeek or Groq via the AI Chat model selector.

3. Failed to fetch / Network Connection Interrupted

• Cause: Browser ad-blockers (e.g. uBlock Origin) or strict VPN firewall settings intercepting outgoing API requests to api.openai.com or openrouter.ai.
• Fix: Whitelist your Eviory domain in your browser extension or disable strict privacy script blockers for the workspace.

Pro-Tip: Multi-Provider Redundancy

To prevent workflow interruptions when one provider experiences downtime:

1. Configure at least two providers in Settings (for instance, OpenRouter as primary and DeepSeek as backup).
2. Use the model dropdown inside AI Chat or Master Workspace to instantly swap engines with zero reconfiguration.