Google Gemini Provider¶

GroundlensGemini wraps the Google Generative AI Python SDK and automatically scores every Gemini response for hallucination risk using groundlens.

Installation¶

pip install "groundlens[google]"

This installs the google-generativeai package.

Quick Start¶

from groundlens.providers.google import GroundlensGemini

llm = GroundlensGemini(api_key="AI...")

# With context (SGI scoring)
resp = llm.chat(
    "What are the main conclusions?",
    context="The report concludes that renewable energy costs have decreased by 70% since 2010...",
)
print(resp.text)
print(resp.groundlens_score.method)         # 'sgi'
print(resp.groundlens_score.flagged)        # False

# Without context (DGI scoring)
resp = llm.chat("How does DNA replication work?")
print(resp.groundlens_score.method)         # 'dgi'

Configuration¶

llm = GroundlensGemini(
    api_key="AI...",
    model="gemini-2.0-flash",             # Gemini model for generation
    groundlens_model="all-MiniLM-L6-v2",    # Sentence-transformer for scoring
    groundlens_threshold=0.45,               # Reserved for future use
)

Parameter	Default	Description
`api_key`	Required	Google AI API key
`model`	`"gemini-2.0-flash"`	Gemini model for generation
`groundlens_model`	`"all-MiniLM-L6-v2"`	Embedding model for scoring
`groundlens_threshold`	`0.45`	Reserved for future threshold customization

Response Object¶

The LLMResponse returned by chat() contains:

Field	Type	Description
`text`	`str`	Generated response text
`model`	`str`	Model identifier
`usage`	`dict`	Token usage (when available from the API)
`groundlens_score`	`GroundlensScore`	Full groundlens evaluation result

Usage metadata includes prompt_token_count, candidates_token_count, and total_token_count when provided by the Gemini API.

Passing Extra Parameters¶

Additional keyword arguments are forwarded to generate_content:

resp = llm.chat(
    "Summarize the research paper.",
    context=paper_abstract,
    generation_config={"temperature": 0.3},
)

Convenience Method¶

resp = llm.complete("Explain the water cycle.", context=source_text)

Environment Variable for API Key¶

import os
from groundlens.providers.google import GroundlensGemini

llm = GroundlensGemini(api_key=os.environ["GOOGLE_API_KEY"])