BerriAI/litellm · error · ValueError

Unable to get Image Response. Please pass a valid llm_provid

Error message

Unable to get Image Response. Please pass a valid llm_provider.

What it means

In image generation, the provider function's result is normalized: a dict is unpacked into ImageResponse, an ImageResponse passes through (cache hits), and a pending coroutine is awaited. If after all that response is still None — the provider returned nothing recognizable — LiteLLM raises ValueError and then re-raises it through exception_type with the provider's error mapping.

Source

Thrown at litellm/images/main.py:118

        # Add the context to the function
        ctx: Final = contextvars.copy_context()
        func_with_context: Final = partial(ctx.run, func)

        _, custom_llm_provider, _, _ = get_llm_provider(model=model, api_base=kwargs.get("api_base", None))

        # Await normally
        init_response: Final = await loop.run_in_executor(None, func_with_context)

        response: ImageResponse | None = None
        if isinstance(init_response, dict):
            response = ImageResponse(**init_response)
        elif isinstance(init_response, ImageResponse):  ## CACHING SCENARIO
            response = init_response
        elif asyncio.iscoroutine(init_response):
            response = await init_response

        if response is None:
            raise ValueError("Unable to get Image Response. Please pass a valid llm_provider.")

        return response
    except Exception as e:
        custom_llm_provider = custom_llm_provider or "openai"
        raise exception_type(
            model=model,
            custom_llm_provider=custom_llm_provider,
            original_exception=e,
            completion_kwargs=args,
            extra_kwargs=kwargs,
        )


# fmt: off

# Overload for when aimg_generation=True (returns Coroutine)
@overload
def image_generation(

View on GitHub (pinned to 6c2dcb801b)

Solutions

Verify image support: use a model/provider with real image support (e.g. openai/dall-e-3, azure image models, gemini image models, stability)
Set custom_llm_provider correctly for your model string instead of omitting/typoing it
Upgrade litellm — image adapters that previously returned None for some paths have been filled in over releases
Reproduce with litellm.image_generation(model='<provider>/<image-model>', prompt=..., custom_llm_provider='<provider>') and inspect the provider call if it still fails

Example fix

# before
r = litellm.image_generation(model='some-oss-model', prompt='a cat')  # provider returns None

# after
r = litellm.image_generation(model='dall-e-3', prompt='a cat', custom_llm_provider='openai')

Defensive patterns

Strategy: fallback

Validate before calling

SUPPORTED_IMAGE_PROVIDERS = {"openai", "azure", "gemini", "stability", "bedrock", "vertex_ai"}
if custom_llm_provider not in SUPPORTED_IMAGE_PROVIDERS:
    raise ValueError(f"{custom_llm_provider} has no image generation support")

Type guard

def provider_supports_images(provider: str) -> bool:
    import litellm
    return provider in getattr(litellm, "image_providers", {"openai", "azure", "gemini", "stability"})

Try / catch

try:
    r = litellm.image_generation(model=m, prompt=p, custom_llm_provider=prov)
except Exception as e:
    if "Unable to get Image Response" in str(e):
        r = litellm.image_generation(model="dall-e-3", prompt=p, custom_llm_provider="openai")  # fallback

Prevention

Always pass custom_llm_provider for image calls
Verify the model actually serves images before wiring it in

When it happens

Trigger: Calling litellm.image_generation with a custom_llm_provider that has no real image implementation (returns None), a provider adapter that silently swallows errors and returns None, or kwargs (e.g. async flags) that route to a code path which never produces a response.

Common situations: Passing an unsupported provider/model for image generation (e.g. a text-only provider); version mismatches where a provider class exists but its image method is a stub; mocking provider calls to return None in tests.

Related errors

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/6f8a59d0b5c1e415. Report an issue: GitHub.