diff --git a/comfy_api/latest/_input_impl/video_types.py b/comfy_api/latest/_input_impl/video_types.py
index ae60603c4..f5af41973 100644
--- a/comfy_api/latest/_input_impl/video_types.py
+++ b/comfy_api/latest/_input_impl/video_types.py
@@ -743,6 +743,8 @@ class VideoFromFile(VideoInput):
                             if to_skip >= resampled.samples:
                                 continue
                             audio_started = True
+                            if duration and frame_start > start_time:
+                                duration_cap = min(duration_cap, math.ceil((start_time + duration - frame_start) * sample_rate))
                             if to_skip:
                                 pending_audio.append(audio_frame_from_ndarray(resampled.to_ndarray()[..., to_skip:]))
                                 continue
diff --git a/tests-unit/comfy_api_test/video_types_test.py b/tests-unit/comfy_api_test/video_types_test.py
index 49b53c6af..ae758bd40 100644
--- a/tests-unit/comfy_api_test/video_types_test.py
+++ b/tests-unit/comfy_api_test/video_types_test.py
@@ -487,6 +487,13 @@ def test_save_to_transcode_learns_unprobed_audio_params():
     assert trimmed_before_audio["audio_codecs"] == []
     assert trimmed_before_audio["audio_seconds"] is None
 
+    buffer.seek(0)
+    trimmed_crossing_audio = transcode_and_probe(VideoFromFile(buffer, start_time=11.5, duration=1))
+    assert trimmed_crossing_audio["frames"] == fps
+    assert trimmed_crossing_audio["audio_codecs"] == ["aac"]
+    assert trimmed_crossing_audio["video_seconds"] == pytest.approx(1.0, abs=0.05)
+    assert trimmed_crossing_audio["audio_seconds"] == pytest.approx(0.5, abs=0.1)
+
 
 def test_save_to_transcode_trimmed_fragmented_mp4_keeps_audio():
     """Fragmented mp4 (MediaRecorder, DASH/HLS-derived files) delivers audio well behind