Gemini 3.1 Flash-Lite: 가장 빠르고 저렴한 Gemini 3 모델

from google.generativeai import GenerativeModel
 
model = GenerativeModel('gemini-3.1-flash-lite')
 
# Level 1: 빠른 번역
response = model.generate_content(
    "Translate to Spanish: Hello world",
    thinking_level=1
)
 
# Level 3: 복잡한 UI 생성
response = model.generate_content(
    "Create a responsive dashboard with...",
    thinking_level=3
)

# 100만 개 문장 번역
# Level 1 사용 → 초당 수백 개 처리
 
texts = load_texts()  # 1M items
for batch in chunks(texts, 1000):
    translate(batch, thinking_level=1)
    
# 비용: 약 $250 (입력만 계산 시)
# 2.5 Flash 대비: 속도 2.5배, 비용 비슷

# Level 1로 빠른 검열
def moderate_content(text):
    response = model.generate_content(
        f"Is this safe? {text}",
        thinking_level=1
    )
    return response.text
 
# 초당 수천 개 처리 가능

# Level 3로 정확한 UI 생성
response = model.generate_content("""
Create a React dashboard with:
- Real-time chart
- Data table with sorting
- Filter sidebar
- Responsive layout
""", thinking_level=3)
 
print(response.text)  # 완전한 React 컴포넌트

# Level 3로 복잡한 로직
response = model.generate_content("""
Generate a traffic simulation with:
- 4-way intersection
- Traffic lights
- Pedestrian crossings
- Emergency vehicles
""", thinking_level=3)

import google.generativeai as genai
 
genai.configure(api_key="YOUR_API_KEY")
model = genai.GenerativeModel('gemini-3.1-flash-lite-preview')
 
response = model.generate_content(
    "Hello world",
    thinking_level=1
)
print(response.text)

from vertexai.preview.generative_models import GenerativeModel
 
model = GenerativeModel('gemini-3.1-flash-lite-preview')
response = model.generate_content("Hello world")