61 lines
2.3 KiB
Python
61 lines
2.3 KiB
Python
import os
|
|
from typing import Any, Optional
|
|
|
|
from langchain_google_genai import ChatGoogleGenerativeAI
|
|
|
|
from .base_client import BaseLLMClient, normalize_content
|
|
from .validators import validate_model
|
|
|
|
|
|
class NormalizedChatGoogleGenerativeAI(ChatGoogleGenerativeAI):
|
|
"""ChatGoogleGenerativeAI with normalized content output.
|
|
|
|
Gemini 3 models return content as list of typed blocks.
|
|
This normalizes to string for consistent downstream handling.
|
|
"""
|
|
|
|
def invoke(self, input, config=None, **kwargs):
|
|
return normalize_content(super().invoke(input, config, **kwargs))
|
|
|
|
|
|
class GoogleClient(BaseLLMClient):
|
|
"""Client for Google Gemini models."""
|
|
|
|
def __init__(self, model: str, base_url: Optional[str] = None, **kwargs):
|
|
super().__init__(model, base_url, **kwargs)
|
|
|
|
def get_llm(self) -> Any:
|
|
"""Return configured ChatGoogleGenerativeAI instance."""
|
|
llm_kwargs = {"model": self.model}
|
|
|
|
for key in ("timeout", "max_retries", "google_api_key", "callbacks", "http_client", "http_async_client"):
|
|
if key in self.kwargs:
|
|
llm_kwargs[key] = self.kwargs[key]
|
|
|
|
# Explicitly fallback to environment variable if google_api_key is missing
|
|
if "google_api_key" not in llm_kwargs:
|
|
env_key = os.getenv("GOOGLE_API_KEY")
|
|
if env_key:
|
|
llm_kwargs["google_api_key"] = env_key
|
|
|
|
# Map thinking_level to appropriate API param based on model
|
|
# Gemini 3 Pro: low, high
|
|
# Gemini 3 Flash: minimal, low, medium, high
|
|
# Gemini 2.5: thinking_budget (0=disable, -1=dynamic)
|
|
thinking_level = self.kwargs.get("thinking_level")
|
|
if thinking_level:
|
|
model_lower = self.model.lower()
|
|
if "gemini-3" in model_lower:
|
|
# Use thinking_budget as Gemini 3 Pro/Flash SDK expects it
|
|
# Mapping: low/minimal -> small budget, high -> large budget
|
|
llm_kwargs["thinking_budget"] = 2000 if thinking_level == "high" else 1000
|
|
else:
|
|
# Gemini 2.5: map to thinking_budget
|
|
llm_kwargs["thinking_budget"] = -1 if thinking_level == "high" else 0
|
|
|
|
return NormalizedChatGoogleGenerativeAI(**llm_kwargs)
|
|
|
|
def validate_model(self) -> bool:
|
|
"""Validate model for Google."""
|
|
return validate_model("google", self.model)
|