feat(qianfan): add official OpenAI-compatible provider support

litellm/constants.py

@@ -778,6 +778,7 @@ openai_compatible_endpoints: List = [
     "https://ai-gateway.vercel.sh/v1",
     "https://api.inference.wandb.ai/v1",
     "https://api.clarifai.com/v2/ext/openai/v1",
+    "https://qianfan.baidubce.com/v2",
 ]
 
 
@@ -814,6 +815,7 @@ openai_compatible_providers: List = [
     "novita",
     "meta_llama",
     "publicai",  # PublicAI - JSON-configured provider
+    "qianfan",  # Qianfan - JSON-configured provider
     "synthetic",  # Synthetic - JSON-configured provider
     "apertis",  # Apertis - JSON-configured provider
     "nano-gpt",  # Nano-GPT - JSON-configured provider

litellm/litellm_core_utils/get_llm_provider_logic.py

@@ -362,6 +362,9 @@ def get_llm_provider(  # noqa: PLR0915
                     elif endpoint == "https://api.inference.wandb.ai/v1":
                         custom_llm_provider = "wandb"
                         dynamic_api_key = get_secret_str("WANDB_API_KEY")
+                    elif endpoint == "https://qianfan.baidubce.com/v2":
+                        custom_llm_provider = "qianfan"
+                        dynamic_api_key = api_key or get_secret_str("QIANFAN_API_KEY")
 
                     if api_base is not None and not isinstance(api_base, str):
                         raise Exception(

litellm/llms/openai_like/providers.json

@@ -11,6 +11,14 @@
       "convert_content_list_to_string": true
     }
   },
+  "qianfan": {
+    "base_url": "https://qianfan.baidubce.com/v2",
+    "api_key_env": "QIANFAN_API_KEY",
+    "api_base_env": "QIANFAN_API_BASE",
+    "param_mappings": {
+      "max_completion_tokens": "max_tokens"
+    }
+  },
   "helicone": {
     "base_url": "https://ai-gateway.helicone.ai/",
     "api_key_env": "HELICONE_API_KEY"

litellm/types/utils.py

@@ -3202,6 +3202,7 @@ class LlmProviders(str, Enum):
     OPENAI = "openai"
     CHATGPT = "chatgpt"
     OPENAI_LIKE = "openai_like"  # embedding only
+    QIANFAN = "qianfan"
     JINA_AI = "jina_ai"
     XAI = "xai"
     ZAI = "zai"

tests/test_litellm/llms/openai_like/test_qianfan.py

@@ -0,0 +1,128 @@
+import os
+import sys
+from unittest.mock import patch
+
+sys.path.insert(0, os.path.abspath("../../../../.."))
+
+import litellm
+from litellm import Router
+from litellm.llms.openai_like.dynamic_config import create_config_class
+from litellm.llms.openai_like.json_loader import JSONProviderRegistry
+from litellm.types.utils import LlmProviders
+from litellm.utils import ProviderConfigManager
+
+
+def test_qianfan_in_provider_list():
+    assert "qianfan" in litellm.provider_list
+    assert "qianfan" in litellm.openai_compatible_providers
+    assert LlmProviders.QIANFAN == "qianfan"
+
+
+def test_qianfan_json_config_exists():
+    config = JSONProviderRegistry.get("qianfan")
+
+    assert config is not None
+    assert config.base_url == "https://qianfan.baidubce.com/v2"
+    assert config.api_key_env == "QIANFAN_API_KEY"
+    assert config.api_base_env == "QIANFAN_API_BASE"
+    assert config.param_mappings["max_completion_tokens"] == "max_tokens"
+
+
+def test_qianfan_dynamic_config_defaults():
+    provider = JSONProviderRegistry.get("qianfan")
+    assert provider is not None
+
+    config = create_config_class(provider)()
+    api_base, api_key = config._get_openai_compatible_provider_info(
+        api_base=None, api_key=None
+    )
+
+    assert api_base == "https://qianfan.baidubce.com/v2"
+    assert api_key is None
+    assert config.custom_llm_provider == "qianfan"
+
+
+def test_qianfan_provider_detection_by_prefix():
+    model, provider, _, api_base = litellm.get_llm_provider(
+        model="qianfan/ernie-4.5-turbo-128k"
+    )
+
+    assert provider == "qianfan"
+    assert model == "ernie-4.5-turbo-128k"
+    assert api_base == "https://qianfan.baidubce.com/v2"
+
+
+def test_qianfan_provider_detection_by_api_base_preserves_explicit_key():
+    model, provider, api_key, api_base = litellm.get_llm_provider(
+        model="ernie-4.5-turbo-128k",
+        api_base="https://qianfan.baidubce.com/v2",
+        api_key="test-qianfan-key",
+    )
+
+    assert model == "ernie-4.5-turbo-128k"
+    assert provider == "qianfan"
+    assert api_key == "test-qianfan-key"
+    assert api_base == "https://qianfan.baidubce.com/v2"
+
+
+def test_qianfan_provider_detection_by_api_base_prefers_explicit_key_over_env():
+    with patch.dict(os.environ, {"QIANFAN_API_KEY": "env-qianfan-key"}):
+        model, provider, api_key, api_base = litellm.get_llm_provider(
+            model="ernie-4.5-turbo-128k",
+            api_base="https://qianfan.baidubce.com/v2",
+            api_key="explicit-qianfan-key",
+        )
+
+    assert model == "ernie-4.5-turbo-128k"
+    assert provider == "qianfan"
+    assert api_key == "explicit-qianfan-key"
+    assert api_base == "https://qianfan.baidubce.com/v2"
+
+
+def test_qianfan_router_config():
+    router = Router(
+        model_list=[
+            {
+                "model_name": "qianfan-router-model",
+                "litellm_params": {
+                    "model": "qianfan/ernie-4.5-turbo-128k",
+                    "api_key": "test-key",
+                },
+            }
+        ]
+    )
+
+    assert len(router.model_list) == 1
+    assert router.model_list[0]["model_name"] == "qianfan-router-model"
+    assert router.deployment_names == ["qianfan/ernie-4.5-turbo-128k"]
+
+
+def test_qianfan_provider_config_manager():
+    config = ProviderConfigManager.get_provider_chat_config(
+        model="ernie-4.5-turbo-128k",
+        provider=LlmProviders.QIANFAN,
+    )
+
+    assert config is not None
+    assert config.custom_llm_provider == "qianfan"
+
+
+def test_qianfan_provider_config_manager_maps_max_completion_tokens():
+    config = ProviderConfigManager.get_provider_chat_config(
+        model="ernie-4.5-turbo-128k",
+        provider=LlmProviders.QIANFAN,
+    )
+
+    mapped_params = config.map_openai_params(
+        non_default_params={
+            "max_completion_tokens": 123,
+            "temperature": 0.7,
+        },
+        optional_params={},
+        model="ernie-4.5-turbo-128k",
+        drop_params=False,
+    )
+
+    assert mapped_params["max_tokens"] == 123
+    assert "max_completion_tokens" not in mapped_params
+    assert mapped_params["temperature"] == 0.7