From 81dd349aea8352bde0d365549a0e4fe913d45c1a Mon Sep 17 00:00:00 2001
From: rootkiller6788 <c8688rickowens@outlook.com>
Date: Thu, 20 Aug 2026 14:21:15 +0800
Subject: [PATCH] fix(python): stop double-counting cached tokens in
 Usage.tally()

Usage.tally() added subusage.total_tokens and then added
subusage.cached_token_count again on top. cached_token_count is already
included in total_tokens (cached tokens are part of the prompt), so the
aggregate total was inflated by the cached token count on every tally.

Remove the redundant addition and add unit tests covering single and
multiple usage accumulation.
---
 sec-gemini-python/sec_gemini/models/usage.py |  1 -
 sec-gemini-python/tests/test_usage.py        | 62 ++++++++++++++++++++
 2 files changed, 62 insertions(+), 1 deletion(-)
 create mode 100644 sec-gemini-python/tests/test_usage.py

diff --git a/sec-gemini-python/sec_gemini/models/usage.py b/sec-gemini-python/sec_gemini/models/usage.py
index 799162d..a5a6d67 100644
--- a/sec-gemini-python/sec_gemini/models/usage.py
+++ b/sec-gemini-python/sec_gemini/models/usage.py
@@ -103,7 +103,6 @@ def tally(self, subusage: Usage) -> None:
       self.cached_token_count += subusage.cached_token_count
       self.thoughts_token_count += subusage.thoughts_token_count
       self.tool_use_prompt_token_count += subusage.tool_use_prompt_token_count
-      self.total_tokens += subusage.cached_token_count
 
   def __repr__(self):
     return (
diff --git a/sec-gemini-python/tests/test_usage.py b/sec-gemini-python/tests/test_usage.py
new file mode 100644
index 0000000..c1ff726
--- /dev/null
+++ b/sec-gemini-python/tests/test_usage.py
@@ -0,0 +1,62 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+
+from sec_gemini.models.usage import Usage
+
+
+def test_tally_accumulates_usage_without_double_counting_cached_tokens():
+  aggregate = Usage()
+  subusage = Usage(
+    prompt_tokens=100,
+    generated_tokens=50,
+    total_tokens=150,
+    cached_token_count=30,
+    thoughts_token_count=10,
+    tool_use_prompt_token_count=5,
+  )
+
+  aggregate.tally(subusage)
+
+  assert aggregate.prompt_tokens == 100
+  assert aggregate.generated_tokens == 50
+  assert aggregate.total_tokens == 150
+  assert aggregate.cached_token_count == 30
+  assert aggregate.thoughts_token_count == 10
+  assert aggregate.tool_use_prompt_token_count == 5
+
+
+def test_tally_accumulates_multiple_usages():
+  aggregate = Usage()
+  aggregate.tally(
+    Usage(
+      prompt_tokens=10,
+      generated_tokens=5,
+      total_tokens=15,
+      cached_token_count=2,
+    )
+  )
+  aggregate.tally(
+    Usage(
+      prompt_tokens=20,
+      generated_tokens=10,
+      total_tokens=30,
+      cached_token_count=4,
+    )
+  )
+
+  assert aggregate.prompt_tokens == 30
+  assert aggregate.generated_tokens == 15
+  assert aggregate.total_tokens == 45
+  assert aggregate.cached_token_count == 6