diff --git a/litellm/litellm_core_utils/streaming_handler.py b/litellm/litellm_core_utils/streaming_handler.py
index fa4650aec4f..6e6a0cb3b65 100644
--- a/litellm/litellm_core_utils/streaming_handler.py
+++ b/litellm/litellm_core_utils/streaming_handler.py
@@ -55,6 +55,7 @@ FUNCTION_CALL_ATTRIBUTE: Final = "function_call"
 _SYNC_ITER_EXHAUSTED: Final = object()
 
 _GCHUNK_FIELDS: Final[frozenset] = frozenset(GChunk.__annotations__)
+_GCHUNK_REQUIRED_FIELDS: Final[frozenset] = frozenset(GChunk.__required_keys__)
 _USAGE_COST_HEADER_PROVIDERS: Final[frozenset[str]] = frozenset({LlmProviders.OPENROUTER.value})
 
 
@@ -2459,7 +2460,7 @@ def generic_chunk_has_all_required_fields(chunk: dict) -> bool:
     :param chunk: The dictionary to check.
     :return: True if all required fields are present, False otherwise.
     """
-    return all(key in _GCHUNK_FIELDS for key in chunk)
+    return all(key in chunk for key in _GCHUNK_REQUIRED_FIELDS)
 
 
 def convert_generic_chunk_to_model_response_stream(
diff --git a/tests/unit/litellm_core_utils/test_streaming_handler.py b/tests/unit/litellm_core_utils/test_streaming_handler.py
index 62d8b0e203f..3cc64f1a727 100644
--- a/tests/unit/litellm_core_utils/test_streaming_handler.py
+++ b/tests/unit/litellm_core_utils/test_streaming_handler.py
@@ -17,6 +17,8 @@ from litellm.litellm_core_utils.streaming_handler import (
     CustomStreamWrapper,
     _ProviderChunkEarlyReturn,
     _ProviderChunkParsed,
+    convert_generic_chunk_to_model_response_stream,
+    generic_chunk_has_all_required_fields,
 )
 from litellm.types.utils import (
     CompletionTokensDetailsWrapper,
@@ -4983,3 +4985,21 @@ async def test_async_stream_without_usage_counts_tokens_off_the_event_loop():
     assert chunks[-1].usage.prompt_tokens > 100_000
     assert chunks[-1].usage.completion_tokens > 100_000
     assert_loop_stayed_free(took, lags)
+
+
+def test_generic_chunk_predicate_rejects_partial_chunk():
+    partial_chunk = {"is_finished": False}
+    assert generic_chunk_has_all_required_fields(partial_chunk) is False
+
+    full_chunk = {
+        "text": "hello",
+        "is_finished": True,
+        "finish_reason": "stop",
+        "usage": None,
+        "index": 0,
+    }
+    assert generic_chunk_has_all_required_fields(full_chunk) is True
+
+    stream = convert_generic_chunk_to_model_response_stream(full_chunk)
+    assert stream.choices[0].delta.content == "hello"
+    assert stream.finish_reason == "stop"