BerriAI/litellm · error · ValueError

api_base not set for LiteLLM Proxy route. Set in env via `LI

Error message

api_base not set for LiteLLM Proxy route. Set in env via `LITELLM_PROXY_API_BASE`

What it means

The image_edit transformation for the litellm_proxy provider builds the URL {api_base}/images/edits for forwarding image edits to a downstream LiteLLM Proxy. api_base is resolved from the parameter or LITELLM_PROXY_API_BASE; when both are None, get_complete_url raises ValueError before any request is sent.

Source

Thrown at litellm/llms/litellm_proxy/image_edit/transformation.py:23

class LiteLLMProxyImageEditConfig(OpenAIImageEditConfig):
    """Configuration for image edit requests routed through LiteLLM Proxy."""

    def validate_environment(
        self,
        headers: dict,
        model: str,
        api_key: str | None = None,
        litellm_params: dict | None = None,
        api_base: str | None = None,
    ) -> dict:
        api_key = api_key or get_secret_str("LITELLM_PROXY_API_KEY")
        headers.update({"Authorization": f"Bearer {api_key}"})
        return headers

    def get_complete_url(self, model: str, api_base: str | None, litellm_params: dict) -> str:
        api_base = api_base or get_secret_str("LITELLM_PROXY_API_BASE")
        if api_base is None:
            raise ValueError("api_base not set for LiteLLM Proxy route. Set in env via `LITELLM_PROXY_API_BASE`")
        api_base = api_base.rstrip("/")
        return f"{api_base}/images/edits"

View on GitHub (pinned to 6c2dcb801b)

Solutions

Set LITELLM_PROXY_API_BASE to the downstream proxy base URL
Or pass api_base="http://downstream-proxy:4000" on the image_edit call / model config
Pair it with LITELLM_PROXY_API_KEY (or api_key) so the Authorization header is valid

Example fix

# before
resp = litellm.image_edit(model="litellm_proxy/gpt-image-1", image=..., prompt=...)

# after
resp = litellm.image_edit(
    model="litellm_proxy/gpt-image-1",
    image=..., prompt=...,
    api_base="http://downstream-proxy:4000",
    api_key="sk-upstream",
)

Defensive patterns

Strategy: validation

Validate before calling

import os

def ensure_proxy_base_for_images(api_base: str | None = None) -> str:
    base = (api_base or os.getenv("LITELLM_PROXY_API_BASE", "")).rstrip("/")
    if not base:
        raise ValueError("Downstream proxy base URL required for image edits")
    return base

Try / catch

try:
    resp = litellm.image_edit(model="litellm_proxy/gpt-image-1", image=img, prompt=p, api_base=ensure_proxy_base_for_images())
except ValueError as e:
    if "api_base not set for LiteLLM Proxy" in str(e):
        # config error — provision the env var; no retry will help
        raise

Prevention

Put api_base in the litellm_proxy image model's litellm_params so callers never pass it manually
Validate image-model deployments at config load: litellm_proxy/* entries must resolve a base URL
Keep LITELLM_PROXY_API_BASE set in every environment that routes images through another proxy

When it happens

Trigger: Calling image editing (model="litellm_proxy/<name>") without api_base on the call and without LITELLM_PROXY_API_BASE in the environment.

Common situations: Proxy-to-proxy image editing configured with only an API key, no base URL; env var present in dev shell but missing in the container running the edits.

Related errors

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/e824155bfdebf7d3. Report an issue: GitHub.