BerriAI/litellm · error · NLPCloudError

{completion_response["error"]}

Error message

{completion_response["error"]}

What it means

Raised by litellm's NLPCloud chat transformer when the parsed JSON response body contains an "error" key — NLP Cloud rejected the request and returned a structured error. The message is the upstream error value and the upstream HTTP status code is attached to the NLPCloudError.

Source

Thrown at litellm/llms/nlp_cloud/chat/transformation.py:194

        encoding: Any,
        api_key: str | None = None,
        json_mode: bool | None = None,
    ) -> ModelResponse:
        ## LOGGING
        logging_obj.post_call(
            input=None,
            api_key=api_key,
            original_response=raw_response.text,
            additional_args={"complete_input_dict": request_data},
        )

        ## RESPONSE OBJECT
        try:
            completion_response: Final = raw_response.json()
        except Exception:
            raise NLPCloudError(message=raw_response.text, status_code=raw_response.status_code)
        if "error" in completion_response:
            raise NLPCloudError(
                message=completion_response["error"],
                status_code=raw_response.status_code,
            )
        else:
            try:
                if len(completion_response["generated_text"]) > 0:
                    model_response.choices[0].message.content = completion_response["generated_text"]
            except Exception:
                raise NLPCloudError(
                    message=json.dumps(completion_response),
                    status_code=raw_response.status_code,
                )

        ## CALCULATING USAGE - baseten charges on time, not tokens - have some mapping of cost here.
        prompt_tokens: Final = completion_response["nb_input_tokens"]
        completion_tokens: Final = completion_response["nb_generated_tokens"]

        model_response.created = int(time.time())

View on GitHub (pinned to 6c2dcb801b)

Solutions

1. Read the message text — it is NLP Cloud's own error description (e.g. 'model not found').
2. Confirm the model string after nlp_cloud/ matches an engine available to your account.
3. Check account credits/quotas on the NLP Cloud dashboard.
4. Strip unsupported optional params from the completion call.

Example fix

# before
out = litellm.completion(model="nlp_cloud/wrong-engine-name", messages=msgs)

# after
out = litellm.completion(model="nlp_cloud/finetuned-llama-3-70b", messages=msgs)

ALLOWED_NLPCLOUD_MODELS = {"finetuned-llama-3-70b", "chatdolphin-30b"}  # keep in sync with your account
if model_name not in ALLOWED_NLPCLOUD_MODELS:
    raise ValueError(f"engine {model_name!r} not enabled on your NLP Cloud account")

from litellm.exceptions import APIError
try:
    out = litellm.completion(model=f"nlp_cloud/{engine}", messages=msgs)
except APIError as e:
    if "error" in str(e).lower() and getattr(e, "status_code", None) in (400, 401, 402, 403):
        raise RuntimeError(f"NLP Cloud rejected request: {e}") from e  # do not retry
    raise

When it happens

Trigger: Calling litellm.completion() with an nlp_cloud/* model when NLP Cloud returns {"error": ...}: invalid model/engine name for your account, malformed request payload, insufficient credits, or an auth error reported in-band.

Common situations: Using an engine name not enabled on the account, exhausted GPU credits, or sending parameters NLP Cloud does not accept.

Related errors

• {raw_response.text}
• {json.dumps(completion_response)}
• completion_response["error"]
• Invalid arg. Model cannot be none.
• Model is None and does not exist in passed completion_respon

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/57264fde95896566. Report an issue: GitHub.