HKUDS/DeepTutor · error · LLMConfigError

OpenAI API key is not configured. Set it in Settings > Catal

Error message

OpenAI API key is not configured. Set it in Settings > Catalog, or select a local provider such as Ollama.

What it means

The OpenAI-compatible provider refuses to instantiate for provider_name='openai' targeting the official api.openai.com endpoint with a placeholder key (None, '', 'no-key', 'sk-no-key-required'), since real API calls would just fail with 401 later. LLMConfigError is raised at construction time with actionable copy.

Source

Thrown at deeptutor/services/llm/provider_core/openai_compat_provider.py:150

        self.extra_headers = extra_headers or {}
        self._spec = spec
        self._provider_name = provider_name

        if primary_key and spec and spec.env_key:
            self._setup_env(primary_key, api_base)

        effective_base = api_base or (spec.default_api_base if spec else None) or None
        self._effective_base = effective_base
        endpoint = (effective_base or "").rstrip("/")
        # api_key may be a list (key pool); only the resolved primary key
        # counts for the configured-key check.
        placeholder_key = primary_key in {None, "", "no-key", "sk-no-key-required"}
        if (
            provider_name == "openai"
            and (not endpoint or endpoint == "https://api.openai.com/v1")
            and placeholder_key
        ):
            raise LLMConfigError(
                "OpenAI API key is not configured. Set it in Settings > Catalog, "
                "or select a local provider such as Ollama."
            )
        default_headers: dict[str, str] = {"x-session-affinity": uuid.uuid4().hex}
        if _uses_openrouter(spec, effective_base):
            default_headers.update(_DEFAULT_OPENROUTER_HEADERS)
        if extra_headers:
            default_headers.update(extra_headers)

        self._client = AsyncOpenAI(
            api_key=primary_key or "no-key",
            base_url=effective_base,
            default_headers=default_headers,
            max_retries=0,
            **openai_client_kwargs(),
        )
        self._responses_failures: dict[str, int] = {}
        self._responses_tripped_at: dict[str, float] = {}

View on GitHub (pinned to 3e82f13042)

Solutions

Set a real OpenAI API key in Settings > Catalog (or OPENAI_API_KEY env var).
Or switch the provider to a local one (Ollama) that doesn't need a key.
If using a custom OpenAI-compatible gateway, set its endpoint so this guard doesn't apply.

Example fix

# before
spec = {'provider': 'openai', 'api_key': 'no-key'}
# after
spec = {'provider': 'openai', 'api_key': os.environ['OPENAI_API_KEY']}

Defensive patterns

Strategy: validation

Validate before calling

PLACEHOLDERS = {None, '', 'no-key', 'sk-no-key-required'}

def openai_provider_instantiable(spec: dict) -> bool:
    if spec.get('provider') != 'openai':
        return True
    custom = spec.get('endpoint') not in (None, '', 'https://api.openai.com/v1')
    return custom or spec.get('api_key') not in PLACEHOLDERS

Try / catch

try:
    prov = build_openai_compat_provider(spec)
except LLMConfigError as e:
    if 'API key is not configured' in str(e):
        spec['api_key'] = os.environ['OPENAI_API_KEY']; prov = build_openai_compat_provider(spec)
    else:
        raise

Prevention

Export OPENAI_API_KEY in the environment or settings before building providers
Fail startup fast when the active provider is OpenAI without a key
Default new setups to a local provider like Ollama when no key exists

When it happens

Trigger: Building the provider with spec for 'openai' with no custom endpoint (or the default https://api.openai.com/v1) and a placeholder key — typical when a local-style keyless config is reused for OpenAI.

Common situations: Switching a local/Ollama setup to OpenAI without adding a key; a UI/settings template that pre-fills 'no-key'; env var OPENAI_API_KEY unset so the pool falls back to a placeholder.

Understand the failure class

Background: "API key is required" / "API key not found" / "No API key was set": the missing-api-key error family across 16 libraries — this error's family across 16 libraries.

Related errors

AI-assisted analysis of HKUDS/DeepTutor@3e82f13042 (2026-08-27). Data as JSON: /api/errors/187cd908d2baac8c. Report an issue: GitHub.