chroma-core/chroma · error · ValueError

The {self.api_key_env_var} environment variable is not set.

Error message

The {self.api_key_env_var} environment variable is not set.

What it means

The legacy GoogleGenerativeAiEmbeddingFunction resolves its credential in priority order: explicit api_key argument (deprecated, emits DeprecationWarning), then GOOGLE_API_KEY if that env var is set (it then takes over api_key_env_var), then the variable named by api_key_env_var (default GEMINI_API_KEY). If none yields a key, initialization fails with this message naming the variable that was finally checked.

Source

Thrown at chromadb/utils/embedding_functions/google_embedding_function.py:258

        except ImportError:
            raise ValueError(
                "The Google Generative AI python package is not installed. Please install it with `pip install google-generativeai`"
            )

        if api_key is not None:
            warnings.warn(
                "Direct api_key configuration will not be persisted. "
                "Please use environment variables via api_key_env_var for persistent storage.",
                DeprecationWarning,
            )
        if os.getenv("GOOGLE_API_KEY") is not None:
            self.api_key_env_var = "GOOGLE_API_KEY"
        else:
            self.api_key_env_var = api_key_env_var

        self.api_key = api_key or os.getenv(self.api_key_env_var)
        if not self.api_key:
            raise ValueError(
                f"The {self.api_key_env_var} environment variable is not set."
            )

        self.model_name = model_name
        self.task_type = task_type
        self.dimension = dimension

        genai.configure(
            api_key=self.api_key,
            client_options={"headers": {"x-goog-api-client": f"chroma/{__version__}"}},
        )
        self._genai = genai

    def __call__(self, input: Documents) -> Embeddings:
        """
        Generate embeddings for the given documents.

        Args:

View on GitHub (pinned to aecdd12c8a)

Solutions

  1. Export one of the recognized variables: export GEMINI_API_KEY=... (or GOOGLE_API_KEY, which takes precedence)
  2. Pass api_key_env_var pointing at a variable your platform actually injects
  3. For quick scripts only: pass api_key='...' (deprecated - not persisted in config)
  4. Migrate to GoogleGeminiEmbeddingFunction, the supported class

Example fix

# before
from chromadb.utils.embedding_functions import GoogleGenerativeAiEmbeddingFunction
ef = GoogleGenerativeAiEmbeddingFunction()  # ValueError: The GEMINI_API_KEY environment variable is not set.

# after
import os
from dotenv import load_dotenv
load_dotenv()  # .env has GEMINI_API_KEY=...
ef = GoogleGenerativeAiEmbeddingFunction()
Defensive patterns

Strategy: validation

Validate before calling

import os
from dotenv import load_dotenv

load_dotenv()
api_key_env = "GOOGLE_API_KEY" if os.getenv("GOOGLE_API_KEY") else "GEMINI_API_KEY"
assert os.getenv(api_key_env), f"Set {api_key_env} before starting"

from chromadb.utils.embedding_functions import GoogleGenerativeAiEmbeddingFunction
ef = GoogleGenerativeAiEmbeddingFunction(api_key_env_var=api_key_env)

Try / catch

try:
    ef = GoogleGenerativeAiEmbeddingFunction()
except ValueError as e:
    if "environment variable is not set" in str(e):
        raise SystemExit(f"Embedding auth missing: {e}") from e
    raise

Prevention

When it happens

Trigger: Constructing with neither GOOGLE_API_KEY nor GEMINI_API_KEY exported and no api_key argument; passing api_key_env_var='MY_KEY' when MY_KEY is unset while GOOGLE_API_KEY is also absent; env present in the login shell but missing in the actual process (docker exec, cron, systemd, notebook kernel started earlier).

Common situations: Local-vs-deployed env drift; secrets not mounted into containers; .env not loaded via python-dotenv; accidentally unsetting GOOGLE_API_KEY that a shared module relied on.

Related errors


AI-assisted analysis of chroma-core/chroma@aecdd12c8a (2026-08-16). Data as JSON: /api/errors/daa3abe66b837cf0. Report an issue: GitHub.