added support for gemini

simplemind/providers/gemini.py

@@ -0,0 +1,71 @@
+import instructor
+import google.generativeai as genai
+
+from ._base import BaseProvider
+from ..settings import settings
+from ..models import Message, Conversation
+
+PROVIDER_NAME = "gemini"
+DEFAULT_MODEL = "models/gemini-1.5-flash-latest"
+
+
+class Gemini(BaseProvider):
+    NAME = PROVIDER_NAME
+    DEFAULT_MODEL = DEFAULT_MODEL
+
+    def __init__(self, api_key: str | None = None):
+        self.api_key = api_key or settings.get_api_key(PROVIDER_NAME)
+        self.model_name = DEFAULT_MODEL
+
+    @property
+    def client(self, model_name: str = DEFAULT_MODEL):
+        """The raw Gemini client."""
+        if not self.api_key:
+            raise ValueError("Gemini API key is required")
+        self.model_name = model_name
+        return genai.GenerativeModel(model_name=model_name)
+
+    @property
+    def structured_client(self):
+        """A Gemini client patched with Instructor."""
+        return instructor.from_gemini(self.client)
+
+    def send_conversation(self, conversation: "Conversation") -> "Message":
+        """Send a conversation to the Gemini API."""
+
+        messages = [
+            {
+                "role": msg.role,
+                "content": msg.text,
+                "metadata": msg.meta or {},
+            }
+            for msg in conversation.messages
+        ]
+
+        response = self.structured_client.chat.completions.create(
+            messages=messages, response_model=None
+        )
+
+        # Create and return a properly formatted Message instance
+        return Message(
+            role="assistant",
+            text=str(response),
+            raw=response,
+            llm_model=self.model_name,
+            llm_provider=PROVIDER_NAME,
+        )
+
+    def structured_response(self, prompt: str, response_model, **kwargs):
+        """Send a structured response to the Gemini API."""
+        response = self.structured_client.chat.completions.create(
+            messages=[{"role": "user", "content": prompt}],
+            response_model=response_model,
+        )
+        return response
+
+    def generate_text(self, prompt: str, **kwargs) -> str:
+        """Generate text using the Gemini API."""
+        response = self.structured_client.chat.completions.create(
+            messages=[{"role": "user", "content": prompt}], response_model=None
+        )
+        return str(response)