ATH-MaaS/Pixelle-Video · error · RuntimeError

OPENAI_API_KEY not set. Configure OpenAI only when using GPT

Error message

OPENAI_API_KEY not set. Configure OpenAI only when using GPT image models.

What it means

The GPT image client (ImageGPT) is created lazily via the gpt_client property and requires OPENAI_API_KEY. If the key is missing when a GPT/OpenAI image model is used, a RuntimeError is raised rather than calling the API unauthenticated.

Source

Thrown at pixelle_video/services/api_services/image_client.py:84

    @property
    def seedream_client(self):
        """Create Seedream client only when a Seedream/ARK model is selected."""
        if not self._ark_api_key:
            raise RuntimeError("ARK_API_KEY not set. Configure ARK only when using Seedream image models.")
        if self._seedream_client is None:
            self._seedream_client = SeedreamClient(
                api_key=self._ark_api_key,
                base_url=self._ark_base_url,
                local_proxy=self._ark_local_proxy,
            )
        return self._seedream_client

    @property
    def gpt_client(self):
        """Create OpenAI image client only when a GPT/OpenAI image model is selected."""
        if not self._gpt_api_key:
            raise RuntimeError("OPENAI_API_KEY not set. Configure OpenAI only when using GPT image models.")
        if self._gpt_client is None:
            self._gpt_client = ImageGPT(
                api_key=self._gpt_api_key,
                base_url=self._gpt_base_url,
                local_proxy=self._gpt_local_proxy,
            )
        return self._gpt_client

    def generate_image(self,
                       prompt: str,
                       image_paths: Optional[List[str]] = None,
                       model: str = "wan2.7-image",
                       save_dir: Optional[str] = None,
                       session_id: Optional[str] = None,
                       video_ratio: Optional[str] = "16:9",
                       resolution: Optional[str] = "2K") -> List[str]:
        """
        Generate images based on prompt and optional reference images.

View on GitHub (pinned to 848b054e4f)

Solutions

Set OPENAI_API_KEY in the environment before generation
Confirm the .env/secret is actually loaded into the process
Check the config layer reads OPENAI_API_KEY into _gpt_api_key

Example fix

// before
export ARK_API_KEY=...
// after
export ARK_API_KEY=...
export OPENAI_API_KEY=sk-...

Defensive patterns

Strategy: validation

Validate before calling

import os
if not os.environ.get("OPENAI_API_KEY"):
    raise SystemExit("Set OPENAI_API_KEY before using GPT image models")

Try / catch

try:
    paths = gen.generate_image(prompt=p)
except RuntimeError as e:
    if "OPENAI_API_KEY not set" in str(e):
        raise SystemExit("Configure OPENAI_API_KEY in your environment") from e

Prevention

Fail fast at boot if required keys are missing
Mount secrets in CI/containers explicitly
Distinguish which provider key each model family needs

When it happens

Trigger: Selecting a GPT image model without OPENAI_API_KEY set; key defined in a different env var; config not reloaded after adding the key.

Common situations: Fresh CI/container without secrets mounted; .env not loaded; typos in the variable name; key revoked and removed from environment.

Understand the failure class

Background: "environment variable is not set" and "Missing keys in environment" errors: what missing required env var messages mean and how to fix them — this error's family across 28 libraries.

Related errors

AI-assisted analysis of ATH-MaaS/Pixelle-Video@848b054e4f (2026-08-30). Data as JSON: /api/errors/3a33a2309e6a6110. Report an issue: GitHub.