replace o3-mini with o4-mini

Author: rishsriv

README.md

@@ -31,7 +31,7 @@ docker compose up --build
 
 We use a simple AI agent with tool use. An LLM attempts to answer a user question with 3 tools – `text_to_sql`, `web_search`, and `pdf_with_citations`.
 
-The model then recursively asks questions using one of these tools until it is satisfied that it has enough context to answer the users question. By default, we use `o3-mini` for text to SQL, `gemini-2.0-flash` for web search, and `claude-3-7-sonnet` for both PDF analysis and orchestration.
+The model then recursively asks questions using one of these tools until it is satisfied that it has enough context to answer the users question. By default, we use `o4-mini` for text to SQL, `gemini-2.0-flash` for web search, and `claude-3-7-sonnet` for both PDF analysis and orchestration.
 
 <img width="730" alt="image" src="https://github.com/user-attachments/assets/1b719e12-e4ea-4e85-82ee-5ac09f07f27a" />
 

backend/query_routes.py

@@ -29,7 +29,7 @@ async def generate_sql_query_route(request: GenerateSQLQueryRequest):
             previous_context=request.previous_context,
             hard_filters=request.hard_filters,
             num_golden_queries=request.num_golden_queries,
-            model_name=request.model_name or "o3-mini",
+            model_name=request.model_name or "o4-mini",
         )
         if resp is None:
             return JSONResponse(

backend/tests/backend_routes/test_tools.py

@@ -412,7 +412,7 @@ async def ticket_summary_tool(input: TicketSummaryInput):
             json={
                 "token": admin_token,
                 "question": user_question,
-                "model": "o3-mini"
+                "model": "o4-mini"
             },
             headers={"Content-Type": "application/json"},
         )

backend/tools/tool_routes.py

@@ -46,7 +46,7 @@ async def answer_question_from_database_route(
     """
     question = request.question
     db_name = request.db_name
-    model = request.model or "o3-mini"
+    model = request.model or "o4-mini"
     return await generate_report_from_question(
         question=question,
         db_name=db_name,

backend/utils_join_hints.py

@@ -42,7 +42,7 @@ async def infer_join_hints(
     ]
     LOGGER.debug(f"Join hints user prompt: {user_prompt}")
     response = await chat_async(
-        model="o3-mini",
+        model="o4-mini",
         messages=messages,
         max_completion_tokens=16384,
         response_format=JoinHints,

backend/utils_sql.py

@@ -523,7 +523,7 @@ async def generate_sql_query(
     previous_context: list[Dict[str, str]] = None,
     hard_filters: list[HardFilter] = None,
     num_golden_queries: int = 4,
-    model_name: str = "o3-mini",
+    model_name: str = "o4-mini",
 ):
     """
     Generate SQL query for a given question, using an LLM.

backend/utils_table_descriptions.py

@@ -156,7 +156,7 @@ async def infer_table_descriptions(
         {"role": "user", "content": user_prompt},
     ]
     response = await chat_async(
-        model="o3-mini",
+        model="o4-mini",
         messages=messages,
         response_format=TableDescriptions,
     )