lmux-google

Google (Gemini) provider for lmux. Talks to the Google Gemini REST API directly over httpx, using google-auth to resolve Vertex AI credentials. Serves Google's models through either backend:

• Vertex AI (default) — authenticated with Google Cloud credentials
• Gemini Developer API (AI Studio) — authenticated with an API key (vertexai=False)

Supports chat completions, streaming, and embeddings for Google-published models: Gemini and the Gemini/text embedding models.

Part of the lmux ecosystem: standardized interface, cost tracking on every response, and registry-based routing across providers.

Auth

Three authentication methods:

Application Default Credentials (default)

Uses google.auth.default(), which works with GOOGLE_APPLICATION_CREDENTIALS, gcloud CLI, or instance metadata.

from lmux_google import GoogleProvider

provider = GoogleProvider(project="my-project", location="us-central1")

Service Account

from lmux_google import GoogleServiceAccountAuthProvider

provider = GoogleProvider(
    project="my-project",
    location="us-central1",
    auth=GoogleServiceAccountAuthProvider(service_account_file="/path/to/key.json"),
)

API Key

Set GOOGLE_API_KEY in your environment:

from lmux_google import GoogleAPIKeyAuthProvider

provider = GoogleProvider(auth=GoogleAPIKeyAuthProvider(), vertexai=False)

Usage

Chat

from lmux import UserMessage

response = provider.chat("gemini-2.5-pro", [UserMessage(content="Hello")])
print(response.content)
print(response.cost)

Streaming

for chunk in provider.chat_stream("gemini-2.5-pro", [UserMessage(content="Hello")]):
    if chunk.delta:
        print(chunk.delta, end="")

Embeddings

response = provider.embed("text-embedding-005", "Hello")
print(response.embeddings)

Async

All methods have async variants: achat, achat_stream, aembed.

Registry

Use with the lmux registry to route across multiple providers:

from lmux import Registry

registry = Registry()
registry.register("google", provider)
response = registry.chat("google/gemini-2.5-pro", messages)

Provider Params

from lmux_google import GoogleParams

response = provider.chat(
    "gemini-2.5-pro",
    messages,
    provider_params=GoogleParams(thinking_config={"thinking_budget": 1024}),
)

Constructor Options

GoogleProvider(
    auth=...,       # AuthProvider, default: GoogleADCAuthProvider()
    project=...,    # GCP project ID
    location=...,   # GCP region
    vertexai=...,   # Use Vertex AI (default: True) vs. AI Studio
    timeout=...,    # request timeout in seconds
    max_retries=..., # retry count for transient failures
    transport=...,        # Optional httpx.BaseTransport for the sync client (proxies, testing)
    async_transport=...,  # Optional httpx.AsyncBaseTransport for the async client
)