Merge pull request #187 from veithly/fix/fix-env

Update Python version compatibility in imports and adjust max_tokens for Gemini model configuration. Refine formatting in __init__.py for improved readability.

Author: moonpyt

spoon_ai/__init__.py

@@ -2,7 +2,7 @@
 
 from pathlib import Path
 
-try:  # Python 3.11+
+try:  # Python 3.12+
     from importlib.metadata import PackageNotFoundError, version as _dist_version
 except Exception:  # pragma: no cover - fallback for older environments
     try:
@@ -26,7 +26,7 @@ def _read_local_pyproject_version() -> str | None:
 
     try:
         try:
-            import tomllib  # Python 3.11+
+            import tomllib  # Python 3.12+
         except Exception:  # pragma: no cover
             import tomli as tomllib  # type: ignore
 
@@ -61,13 +61,13 @@ def _resolve_version() -> str:
 
 __version__: str = _resolve_version()
 
-from spoon_ai.chat import ChatBot 
-from spoon_ai.schema import LLMResponse, LLMResponseChunk, Message 
+from spoon_ai.chat import ChatBot
+from spoon_ai.schema import LLMResponse, LLMResponseChunk, Message
 
 __all__ = [
     "__version__",
     "ChatBot",
-    "Message", 
+    "Message",
     "LLMResponse",
     "LLMResponseChunk",
 ]

spoon_ai/llm/config.py

@@ -341,7 +341,7 @@ def _get_provider_defaults(self, provider_name: str) -> Dict[str, Any]:
             },
             'gemini': {
                 'model': 'gemini-2.5-pro',
-                'max_tokens': 250000,
+                'max_tokens': 20000,
                 'base_url': 'https://generativelanguage.googleapis.com/v1beta',
                 'temperature': 0.1,   # Lower temperature for Gemini
                 **{k: v for k, v in common_defaults.items() if k != 'temperature'}

spoon_ai/llm/providers/gemini_provider.py

@@ -569,6 +569,7 @@ def get_metadata(self) -> ProviderMetadata:
                 ProviderCapability.IMAGE_GENERATION,
                 ProviderCapability.VISION
             ],
+            max_tokens=self.max_tokens,
             supports_system_messages=True,
             rate_limits={
                 "requests_per_minute": 60,
@@ -615,4 +616,4 @@ async def _handle_error(self, error: Exception) -> None:
         elif "timeout" in error_str or "connection" in error_str:
             raise NetworkError("gemini", "Network error", original_error=error)
         else:
-            raise ProviderError("gemini", f"Request failed: {str(error)}", original_error=error)
+            raise ProviderError("gemini", f"Request failed: {str(error)}", original_error=error)