Merge 6facd4e229 into 44d84360fb

litellm/constants.py

@@ -835,6 +835,7 @@ openai_compatible_providers: Final[list] = [
     "nano-gpt",  # Nano-GPT - JSON-configured provider
     "poe",  # Poe - JSON-configured provider
     "chutes",  # Chutes - JSON-configured provider
+    "byesu",  # Byesu - JSON-configured provider
     "parasail",  # Parasail - JSON-configured provider
     "libertai",  # LibertAI - JSON-configured provider
     "featherless_ai",

litellm/llms/openai_like/providers.json

@@ -58,6 +58,16 @@
       "max_completion_tokens": "max_tokens"
     }
   },
+  "byesu": {
+    "base_url": "https://byesu.com/v1",
+    "api_key_env": "BYESU_API_KEY",
+    "api_base_env": "BYESU_API_BASE",
+    "base_class": "openai_gpt",
+    "param_mappings": {
+      "max_completion_tokens": "max_tokens"
+    },
+    "supported_endpoints": ["/v1/chat/completions", "/v1/responses"]
+  },
   "chutes": {
     "base_url": "https://llm.chutes.ai/v1/",
     "api_key_env": "CHUTES_API_KEY",

litellm/types/utils.py

@@ -3827,6 +3827,7 @@ class LlmProviders(str, Enum):
     NANOGPT = "nano-gpt"
     POE = "poe"
     CHUTES = "chutes"
+    BYESU = "byesu"
     NEOSANTARA = "neosantara"
     PARASAIL = "parasail"
     XIAOMI_MIMO = "xiaomi_mimo"

provider_endpoints_support.json

@@ -2989,6 +2989,22 @@
         "rerank": false,
         "a2a": false
       }
+    },
+    "byesu": {
+      "display_name": "Byesu (`byesu`)",
+      "endpoints": {
+        "chat_completions": true,
+        "messages": false,
+        "responses": true,
+        "embeddings": false,
+        "image_generations": false,
+        "audio_transcriptions": false,
+        "audio_speech": false,
+        "moderations": false,
+        "batches": false,
+        "rerank": false,
+        "a2a": false
+      }
     }
   },
   "endpoints": {

tests/test_litellm/llms/openai_like/test_byesu_provider.py

@@ -0,0 +1,97 @@
+"""
+Tests for the `byesu` JSON-configured OpenAI-compatible provider.
+
+byesu is registered purely via litellm/llms/openai_like/providers.json, so these
+tests exercise the JSON registry, provider routing, and the dynamically generated
+config class. No network calls are made.
+"""
+
+import os
+import sys
+
+try:
+    import pytest  # noqa: F401
+except ImportError:
+    pytest = None  # allow standalone execution
+
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "../../../..")))
+
+import litellm  # noqa: E402
+from litellm import get_llm_provider  # noqa: E402
+
+
+class TestByesuProvider:
+    def test_byesu_loaded_from_json(self):
+        """byesu is discovered by the JSON provider registry."""
+        from litellm.llms.openai_like.json_loader import JSONProviderRegistry
+
+        assert JSONProviderRegistry.exists("byesu")
+
+        cfg = JSONProviderRegistry.get("byesu")
+        assert cfg is not None
+        assert cfg.base_url == "https://byesu.com/v1"
+        assert cfg.api_key_env == "BYESU_API_KEY"
+        assert cfg.api_base_env == "BYESU_API_BASE"
+
+    def test_byesu_in_openai_compatible_providers(self):
+        """byesu is registered as an OpenAI-compatible provider."""
+        from litellm.constants import openai_compatible_providers
+
+        assert "byesu" in openai_compatible_providers
+
+    def test_byesu_get_llm_provider_routing(self):
+        """`byesu/<model>` resolves to custom_llm_provider == 'byesu'."""
+        model, custom_llm_provider, _, _ = get_llm_provider(model="byesu/gpt-4o-mini")
+        assert custom_llm_provider == "byesu"
+        assert model == "gpt-4o-mini"
+
+    def test_byesu_api_base_and_key_resolution(self):
+        """The dynamic config resolves the byesu base URL and honors overrides."""
+        from litellm.llms.openai_like.dynamic_config import create_config_class
+        from litellm.llms.openai_like.json_loader import JSONProviderRegistry
+
+        config_class = create_config_class(JSONProviderRegistry.get("byesu"))
+        config = config_class()
+
+        # Defaults to the byesu base URL
+        api_base, _ = config._get_openai_compatible_provider_info(None, "test-key")
+        assert api_base == "https://byesu.com/v1"
+
+        # Explicit overrides win
+        api_base, api_key = config._get_openai_compatible_provider_info(
+            "https://custom.example.com/v1", "override-key"
+        )
+        assert api_base == "https://custom.example.com/v1"
+        assert api_key == "override-key"
+
+    def test_byesu_param_mapping(self):
+        """max_completion_tokens is mapped to max_tokens per providers.json."""
+        from litellm.llms.openai_like.dynamic_config import create_config_class
+        from litellm.llms.openai_like.json_loader import JSONProviderRegistry
+
+        config_class = create_config_class(JSONProviderRegistry.get("byesu"))
+        config = config_class()
+
+        result = config.map_openai_params(
+            {"max_completion_tokens": 128},
+            {},
+            "gpt-4o-mini",
+            False,
+        )
+        assert result.get("max_tokens") == 128
+        assert "max_completion_tokens" not in result
+
+    def test_byesu_declares_responses_api(self):
+        """byesu declares /v1/responses support."""
+        from litellm.llms.openai_like.json_loader import JSONProviderRegistry
+
+        assert JSONProviderRegistry.supports_responses_api("byesu")
+
+
+if __name__ == "__main__":
+    t = TestByesuProvider()
+    for name in dir(t):
+        if name.startswith("test_"):
+            getattr(t, name)()
+            print(f"PASS {name}")
+    print("All byesu provider tests passed.")