Error Handling

The proxy preserves upstream error types. If the LLM provider returns an error, you get the same exception you'd get without Prysm.

import openai

try:
    response = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": "test"}],
    )
except openai.AuthenticationError:
    print("Invalid upstream API key (check project settings)")
except openai.RateLimitError:
    print("Provider rate limited")
except openai.APIError as e:
    print(f"API error: {e}")

Prysm-Specific Errors

Troubleshooting