BerriAI/litellm · error · Timeout

APITimeoutError - Request timed out. Error_str: {error_str}

Error message

APITimeoutError - Request timed out. Error_str: {error_str}

What it means

LiteLLM scans the string of the original exception for phrases like 'Request Timeout Error', 'Request timed out', 'Timed out generating response', or 'The read operation timed out', and re-raises them as a unified Timeout exception. This normalizes client-side and provider-side timeout wording into one type.

Source

Thrown at litellm/litellm_core_utils/exception_mapping_utils.py:2256

                pass

            ################################################################################
            # End of Common Extra information Needed for all providers
            ################################################################################

            ################################################################################
            #################### Start of Provider Exception mapping ####################
            ################################################################################

            if (
                "Request Timeout Error" in error_str
                or "Request timed out" in error_str
                or "Timed out generating response" in error_str
                or "The read operation timed out" in error_str
            ):
                exception_mapping_worked = True

                raise Timeout(
                    message=f"APITimeoutError - Request timed out. Error_str: {error_str}",
                    model=model,
                    llm_provider=custom_llm_provider,
                    litellm_debug_info=extra_information,
                )

            if (
                custom_llm_provider == "litellm_proxy"
            ):  # handle special case where calling litellm proxy + exception str contains error message
                extract_and_raise_litellm_exception(
                    response=getattr(original_exception, "response", None),
                    error_str=error_str,
                    model=model,
                    custom_llm_provider=custom_llm_provider,
                )
            if (
                custom_llm_provider == "openai"
                or custom_llm_provider == "text-completion-openai"

View on GitHub (pinned to 6c2dcb801b)

Solutions

Raise the timeout parameter on the call (litellm.completion(..., timeout=300)).
Set a model-specific default timeout via Router/model_list litellm_params.timeout.
Enable streaming so partial results arrive before the deadline.
For local models, use faster hardware/quantization or a smaller model.

Example fix

# before
resp = litellm.completion(model='ollama/llama3', messages=msgs)

# after
resp = litellm.completion(model='ollama/llama3', messages=msgs, timeout=600, stream=True)

Defensive patterns

Strategy: retry

Try / catch

try {
  await litellm.completion(...);
} catch (e) {
  if (e instanceof litellm.Timeout) { /* raise per-call timeout or reduce workload */ }
}

Prevention

Set timeout proportional to expected generation length.
Enable streaming for long outputs.
Catch litellm.Timeout specifically and retry idempotent requests.

When it happens

Trigger: Calls where the HTTP client read/connect timeout expires (httpx.ReadTimeout surfaced as 'Request timed out'), or the provider returns a body/timeouts with those exact phrases (e.g. 'Timed out generating response' from streaming backends).

Common situations: Default 600s litellm timeout too short for long generations, low per-call timeout passed, slow local inference hardware (Ollama on CPU), or provider under heavy load.

Understand the failure class

Timeouts: ETIMEDOUT, deadlines, and hung requests — what actually expires when a request times out.

Related errors

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/2b67a513c482dc40. Report an issue: GitHub.