pass llm_provider, model in list,

tabs/run_benchmark.py

@@ -14,7 +14,7 @@ def run_benchmark_main(tool_name, model_name, num_questions, openai_api_key, ant
     kwargs["num_questions"] = num_questions
     kwargs["tools"] = [tool_name]
     if model_name:
-        kwargs["model"] = model_name
+        kwargs["model"] = [model_name]
     kwargs["api_keys"] = {}
     if openai_api_key:
         kwargs["api_keys"]["openai"] = openai_api_key
@@ -22,7 +22,14 @@ def run_benchmark_main(tool_name, model_name, num_questions, openai_api_key, ant
         kwargs["api_keys"]["anthropic"] = anthropic_api_key
     if openrouter_api_key:
         kwargs["api_keys"]["openrouter"] = openrouter_api_key
-        
+
+    if "gpt" in model:
+        kwargs["llm_provider"] = "openai"
+    elif "claude" in model:
+        kwargs["llm_provider"] = "anthropic"
+    else:
+        kwargs["llm_provider"] = "openrouter"
+
     kwargs["num_urls"] = 3
     kwargs["num_words"] = 300
     kwargs["provide_source_links"] = True