diff --git a/litellm/litellm_core_utils/core_helpers.py b/litellm/litellm_core_utils/core_helpers.py
index 7c8e2ebeaff4..4113fa351ae7 100644
--- a/litellm/litellm_core_utils/core_helpers.py
+++ b/litellm/litellm_core_utils/core_helpers.py
@@ -1,11 +1,11 @@
 # What is this?
 ## Helper utilities
-from typing import TYPE_CHECKING, Any, Iterable, List, Literal, Optional, Union
+from typing import TYPE_CHECKING, Any, Iterable, List, Literal, Optional, Union, get_args
 
 import httpx
 
 from litellm._logging import verbose_logger
-from litellm.types.llms.openai import AllMessageValues
+from litellm.types.llms.openai import AllMessageValues, OpenAIChatCompletionFinishReason
 
 if TYPE_CHECKING:
     from opentelemetry.trace import Span as _Span
@@ -58,6 +58,12 @@ def safe_divide(
     return numerator / denominator
 
 
+# Module-level constant derived from the source-of-truth Literal type.
+# Avoids recreating the set on every call (map_finish_reason is called per-chunk
+# during streaming) and stays in sync when the Literal is updated.
+_VALID_OPENAI_FINISH_REASONS = frozenset(get_args(OpenAIChatCompletionFinishReason))
+
+
 def map_finish_reason(
     finish_reason: str,
 ):  # openai supports 5 stop sequences - 'stop', 'length', 'function_call', 'content_filter', 'null'
@@ -96,6 +102,18 @@ def map_finish_reason(
         return "tool_calls"
     elif finish_reason == "compaction":
         return "length"
+    # Unknown finish_reason values (e.g. provider-specific error codes like
+    # "network_error" from ZhipuAI/GLM-5) are not in OpenAIChatCompletionFinishReason
+    # Literal and will cause a Pydantic ValidationError in Choices.__init__.
+    # Map them to "finish_reason_unspecified" so the stream can be assembled
+    # without raising an exception.
+    if finish_reason not in _VALID_OPENAI_FINISH_REASONS:
+        verbose_logger.warning(
+            "litellm.map_finish_reason: unknown finish_reason %r from provider; "
+            "mapping to 'finish_reason_unspecified' to avoid ValidationError.",
+            finish_reason,
+        )
+        return "finish_reason_unspecified"
     return finish_reason
 
 
diff --git a/tests/test_litellm/litellm_core_utils/test_core_helpers.py b/tests/test_litellm/litellm_core_utils/test_core_helpers.py
index cd9c401143e3..4118b4cb058f 100644
--- a/tests/test_litellm/litellm_core_utils/test_core_helpers.py
+++ b/tests/test_litellm/litellm_core_utils/test_core_helpers.py
@@ -1,6 +1,66 @@
 """Tests for litellm_core_utils.core_helpers module."""
 
-from litellm.litellm_core_utils.core_helpers import reconstruct_model_name
+import pytest
+
+from litellm.litellm_core_utils.core_helpers import map_finish_reason, reconstruct_model_name
+
+
+class TestMapFinishReason:
+    @pytest.mark.parametrize(
+        "value",
+        [
+            "stop",
+            "length",
+            "function_call",
+            "tool_calls",
+            "content_filter",
+            "finish_reason_unspecified",
+            "eos",
+            "guardrail_intervened",
+            "malformed_function_call",
+        ],
+    )
+    def test_known_openai_values_pass_through(self, value: str) -> None:
+        assert map_finish_reason(value) == value
+
+    def test_anthropic_tool_use_maps_to_tool_calls(self) -> None:
+        assert map_finish_reason("tool_use") == "tool_calls"
+
+    def test_anthropic_max_tokens_maps_to_length(self) -> None:
+        assert map_finish_reason("max_tokens") == "length"
+
+    def test_anthropic_end_turn_maps_to_stop(self) -> None:
+        assert map_finish_reason("end_turn") == "stop"
+
+    def test_cohere_complete_maps_to_stop(self) -> None:
+        assert map_finish_reason("COMPLETE") == "stop"
+
+    def test_cohere_max_tokens_maps_to_length(self) -> None:
+        assert map_finish_reason("MAX_TOKENS") == "length"
+
+    def test_cohere_error_toxic_maps_to_content_filter(self) -> None:
+        assert map_finish_reason("ERROR_TOXIC") == "content_filter"
+
+    def test_vertex_ai_stop_maps_to_stop(self) -> None:
+        assert map_finish_reason("STOP") == "stop"
+
+    def test_vertex_ai_safety_maps_to_content_filter(self) -> None:
+        assert map_finish_reason("SAFETY") == "content_filter"
+
+    def test_vertex_ai_finish_reason_unspecified_maps_correctly(self) -> None:
+        assert map_finish_reason("FINISH_REASON_UNSPECIFIED") == "finish_reason_unspecified"
+
+    def test_vertex_ai_malformed_function_call_maps_correctly(self) -> None:
+        assert map_finish_reason("MALFORMED_FUNCTION_CALL") == "malformed_function_call"
+
+    def test_unknown_value_maps_to_finish_reason_unspecified(self) -> None:
+        assert map_finish_reason("some_unknown_reason") == "finish_reason_unspecified"
+
+    def test_empty_string_maps_to_finish_reason_unspecified(self) -> None:
+        assert map_finish_reason("") == "finish_reason_unspecified"
+
+    def test_zhipuai_glm_network_error_regression(self) -> None:
+        assert map_finish_reason("network_error") == "finish_reason_unspecified"
 
 
 def test_reconstruct_model_name_prefers_deployment_value():