GeminiCacheAdapter

Wraps google.generativeai.GenerativeModel generate_content calls with response caching. Returns cached results for identical (model, contents, params) combinations.

Installation

pip install 'chengeta-ai[gemini]'

---

Usage

Sync

import google.generativeai as genai
from chengeta_ai import CacheManager, InMemoryBackend, CacheKeyBuilder
from chengeta_ai.adapters.gemini_adapter import GeminiCacheAdapter

genai.configure(api_key="your-api-key")
model = genai.GenerativeModel("gemini-2.0-flash")

manager = CacheManager(
    backend=InMemoryBackend(),
    key_builder=CacheKeyBuilder(namespace="myapp"),
)
adapter = GeminiCacheAdapter(model, manager)

response = adapter.generate_content("What is the capital of France?")

# Second call — cache hit, no API cost
response = adapter.generate_content("What is the capital of France?")

Async

response = await adapter.agenerate_content("Summarise the quarterly report.")

Multi-turn content

contents = [
    {"role": "user", "parts": ["Hello"]},
    {"role": "model", "parts": ["Hi! How can I help?"]},
    {"role": "user", "parts": ["What is 2+2?"]},
]
response = adapter.generate_content(contents)

With Redis shared cache

from chengeta_ai.backends.redis_backend import RedisBackend

manager = CacheManager(
    backend=RedisBackend(url="redis://localhost:6379/0"),
    key_builder=CacheKeyBuilder(namespace="gemini"),
)
adapter = GeminiCacheAdapter(model, manager)

Invalidate model entries

from chengeta_ai.core.invalidation import InvalidationEngine

manager = CacheManager(
    backend=InMemoryBackend(),
    key_builder=CacheKeyBuilder(),
    invalidation_engine=InvalidationEngine(InMemoryBackend()),
)
adapter = GeminiCacheAdapter(model, manager)
adapter.invalidate_model()

Transparent proxy

Any attribute not overridden (e.g. model.model_name, model.tools) is forwarded to the underlying GenerativeModel via __getattr__.

---

API Reference

Method	Description
generate_content(contents, **kwargs)	Cached model.generate_content() (sync)
agenerate_content(contents, **kwargs)	Cached model.generate_content_async() (async)
invalidate_model()	Invalidate all cached responses for this model