moving 4o/4omini to 4.1/4.1mini and 4x output tokens

analysis/get_clustering_csv.py

@@ -37,7 +37,7 @@ def wrapper(*args, **kwargs):
         }
     return wrapper
 
-def openai_json_response(messages, model="gpt-4o-mini", temp=1, max_tokens=1024):
+def openai_json_response(messages, model="gpt-4.1", temp=1, max_tokens=4096):
     response = openai_client.chat.completions.create(
         model=model,
         messages=messages,

pipelines/email/preprocess.py

@@ -8,7 +8,7 @@
 
 openai_client = OpenAI()
 
-def openai_json_response(messages, model="gpt-4o-mini", temp=1, max_tokens=1024):
+def openai_json_response(messages, model="gpt-4.1", temp=1, max_tokens=4096):
     response = openai_client.chat.completions.create(
         model=model,
         messages=messages,

server/tools.py

@@ -177,7 +177,7 @@ def choose_tool_and_rewrite(tools, memory, query_text):
     response = openai_json_response([
         tool_and_rewrite(tools, memory),
         user_query(query_text),
-    ], model="gpt-4o")
+    ], model="gpt-4.1")
     tool: Tool | None = response["tool"]
     query_rewrite = response["query_rewrite"]
     return tool, query_rewrite

server/utils.py

@@ -57,7 +57,7 @@ def wrapper(*args, **kwargs):
         }
     return wrapper
 
-def openai_json_response(messages: List, model="gpt-4o-mini", temp=1, max_tokens=1024):
+def openai_json_response(messages: List, model="gpt-4.1", temp=1, max_tokens=4096):
     response = openai_client.chat.completions.create(
         model=model,
         messages=messages,
@@ -72,7 +72,7 @@ def openai_json_response(messages: List, model="gpt-4o-mini", temp=1, max_tokens
     )
     return json.loads(response.choices[0].message.content)
 
-def openai_stream(messages: List, model="gpt-4o-mini", temp=1, max_tokens=1024):
+def openai_stream(messages: List, model="gpt-4.1-mini", temp=1, max_tokens=1024):
     stream = openai_client.chat.completions.create(
         model=model,
         messages=messages,