Add vivo BlueLM-3.6-35B-A3B-0806 to leaderboard

berkeley-function-call-leaderboard/bfcl_eval/constants/model_config.py

@@ -1,6 +1,6 @@
 from dataclasses import dataclass
 from typing import Optional
-
+from bfcl_eval.model_handler.api_inference.vivo import VivoAPIHandler
 from bfcl_eval.model_handler.api_inference.claude import ClaudeHandler
 from bfcl_eval.model_handler.api_inference.cohere import CohereHandler
 from bfcl_eval.model_handler.api_inference.deepseek import DeepSeekAPIHandler
@@ -117,6 +117,18 @@ class ModelConfig:
 
 # Inference through API calls
 api_inference_model_map = {
+    "BlueLM-3.6-35B-A3B-0806": ModelConfig(
+    model_name="BlueLM-3.6-35B-A3B-0806",
+    display_name="BlueLM-3.6-35B-A3B-0806",
+    url="https://api-ai.vivo.com.cn/v1/completions",
+    org="vivo",
+    license="Proprietary",
+    model_handler=VivoAPIHandler,
+    input_price=None,
+    output_price=None,
+    is_fc_model=True,
+    underscore_to_dot=True,
+    ),
     "gorilla-openfunctions-v2": ModelConfig(
         model_name="gorilla-openfunctions-v2",
         display_name="Gorilla-OpenFunctions-v2 (FC)",

berkeley-function-call-leaderboard/bfcl_eval/constants/supported_models.py

@@ -187,4 +187,5 @@
     "qwen3-4b-think-FC",
     "qwen3-4b-nothink-FC",
     "DM-Cito-32B-v1",
+    "BlueLM-3.6-35B-A3B-0806",
 ]

berkeley-function-call-leaderboard/bfcl_eval/model_handler/api_inference/vivo.py

@@ -0,0 +1,59 @@
+import json
+import os
+
+from bfcl_eval.model_handler.api_inference.openai_completion import OpenAICompletionsHandler
+from openai import OpenAI
+from overrides import override
+from typing import Any
+from bfcl_eval.model_handler.model_style import ModelStyle
+
+
+class VivoAPIHandler(OpenAICompletionsHandler):
+    """
+    Handler for vivo BlueLM models accessed via the vivo public API.
+    The API is OpenAI-compatible and supports function calling.
+    """
+
+    def __init__(self, model_name, temperature) -> None:
+        super().__init__(model_name, temperature)
+        self.model_style = ModelStyle.OpenAI
+        self.client = OpenAI(
+            api_key=os.getenv("VIVO_API_KEY", ""),
+            base_url="https://api-ai.vivo.com.cn/v1"
+        )
+        self.is_fc_model = True
+
+    @override
+    def _parse_query_response_FC(self, api_response: Any) -> dict:
+        try:
+            model_responses = [
+                {func_call.function.name: func_call.function.arguments}
+                for func_call in api_response.choices[0].message.tool_calls
+            ]
+            tool_call_ids = [
+                func_call.id for func_call in api_response.choices[0].message.tool_calls
+            ]
+            if len(model_responses) == 0:
+                model_responses = api_response.choices[0].message.content
+        except:
+            model_responses = api_response.choices[0].message.content
+            tool_call_ids = []
+
+        model_responses_message_for_chat_history = api_response.choices[0].message.__dict__
+        if model_responses_message_for_chat_history.get("tool_calls", None):
+            model_responses_message_for_chat_history["tool_calls"] = [
+                {
+                    "id": tool_call.id,
+                    "type": "function",
+                    "function": tool_call.function.__dict__,
+                }
+                for tool_call in model_responses_message_for_chat_history["tool_calls"]
+            ]
+
+        return {
+            "model_responses": model_responses,
+            "model_responses_message_for_chat_history": model_responses_message_for_chat_history,
+            "tool_call_ids": tool_call_ids,
+            "input_token": api_response.usage.prompt_tokens,
+            "output_token": api_response.usage.completion_tokens,
+        }