Use response.set_usage(), closes #29

Refs https://github.com/simonw/llm/issues/610

llm_claude_3.py

@@ -231,6 +231,13 @@ class _Shared:
             kwargs["extra_headers"] = self.extra_headers
         return kwargs
 
+    def set_usage(self, response):
+        usage = response.response_json.pop("usage")
+        if usage:
+            response.set_usage(
+                input=usage.get("input_tokens"), output=usage.get("output_tokens")
+            )
+
     def __str__(self):
         return "Anthropic Messages: {}".format(self.model_id)
 
@@ -250,6 +257,7 @@ class ClaudeMessages(_Shared, llm.Model):
             completion = client.messages.create(**kwargs)
             yield completion.content[0].text
             response.response_json = completion.model_dump()
+        self.set_usage(response)
 
 
 class ClaudeMessagesLong(ClaudeMessages):
@@ -270,6 +278,7 @@ class AsyncClaudeMessages(_Shared, llm.AsyncModel):
             completion = await client.messages.create(**kwargs)
             yield completion.content[0].text
             response.response_json = completion.model_dump()
+        self.set_usage(response)
 
 
 class AsyncClaudeMessagesLong(AsyncClaudeMessages):

pyproject.toml

@@ -9,7 +9,7 @@ classifiers = [
     "License :: OSI Approved :: Apache Software License"
 ]
 dependencies = [
-    "llm>=0.18",
+    "llm>=0.19a0",
     "anthropic>=0.39.0",
 ]
 

tests/test_claude_3.py

@@ -30,8 +30,10 @@ def test_prompt():
         "stop_reason": "end_turn",
         "stop_sequence": None,
         "type": "message",
-        "usage": {"input_tokens": 17, "output_tokens": 15},
     }
+    assert response.input_tokens == 17
+    assert response.output_tokens == 15
+    assert response.token_details is None
 
 
 @pytest.mark.vcr
@@ -50,8 +52,10 @@ async def test_async_prompt():
         "stop_reason": "end_turn",
         "stop_sequence": None,
         "type": "message",
-        "usage": {"input_tokens": 17, "output_tokens": 15},
     }
+    assert response.input_tokens == 17
+    assert response.output_tokens == 15
+    assert response.token_details is None
 
 
 EXPECTED_IMAGE_TEXT = (
@@ -86,5 +90,7 @@ def test_image_prompt():
         "stop_reason": "end_turn",
         "stop_sequence": None,
         "type": "message",
-        "usage": {"input_tokens": 76, "output_tokens": 75},
     }
+    assert response.input_tokens == 76
+    assert response.output_tokens == 75
+    assert response.token_details is None