Error Type

Description

MODEL_LOADING_FAILED

The model failed to load correctly on the node (missing weights, incompatible environment).

INFERENCE_TIMEOUT

The model did not return a response in time (timeout exceeded).

OUT_OF_MEMORY

GPU or CPU memory exhausted during model execution.

INVALID_INPUT

The input data was malformed or incompatible with the model’s schema.

MODEL_OVERLOADED

The model endpoint is receiving too many concurrent requests; try again later.

GENERATION_FAILED

Inference or generation process crashed or produced invalid output.

CONTEXT_LENGTH_EXCEEDED

Prompt or input tokens exceeded the model’s context window.

MODEL_NOT_FOUND

Requested model name or version does not exist or is unavailable.

INTERNAL_ENGINE_ERROR

Generic backend execution error (engine failure, dependency issue).

NETWORK_ERROR

Communication issue between Chutes node and backend compute.

RATE_LIMIT_HIT

SDK-side throttle triggered before sending the request to the API.

NO_ASSISTANT_MESSAGE

API responded without valid assistant output (empty model response).

INVALID_TOKEN

API key recognized but not authorized for this environment or account.