localcloud is cloud for agents for developers, platform/DevOps engineers and teams running CI, plus AI coding agent users. Individuals, teams, nonprofits, and companies can build and test with 25 Google Cloud services on localhost in one Docker container, without a GCP account, credentials or billing. Free during the public preview.
LocalCloud runs Google Cloud-compatible services locally for development, integration tests, and CI. Connect your existing SDKs through local endpoints, or use the LocalCloud MCP server. No Google Cloud account or credentials are required for local workflows.
runtime
one Docker container
services
25 local service APIs
connect
Google Cloud SDKs / stdio MCP
scope
development / tests / CI
localcloud is cloud for agents for developers, platform/DevOps engineers and teams running CI, plus AI coding agent users. Individuals, teams, nonprofits, and companies can build and test with 25 Google Cloud services on localhost in one Docker container, without a GCP account, credentials or billing. Free during the public preview.
$0 costNo setupFast developmentlocalcloud is cloud for agents.
localcloud is cloud for agents for developers, platform/DevOps engineers and teams running CI, plus AI coding agent users. Individuals, teams, nonprofits, and companies can build and test with 25 Google Cloud services on localhost in one Docker container, without a GCP account, credentials or billing. Free during the public preview.
Free for OSS developers. No account or license key required. OSS terms
No account or license key. Free during the public preview, and free for open-source projects after it. CLI releases are Sigstore-signed with SHA-256 checksums.
No cloud accountNo GCP billingBuild and test locally
25
Google Cloud services
each with documented compatibility limits
1
Docker container
persistent data on localhost ports the CLI picks
Choose your next step
Choose services — Find service purposes, exact IDs, startup requirements, and client examples.
Connect through MCP — Use the LocalCloud CLI as a stdio MCP server in a coding client.
Start with repository context
Inspect the repository and its existing agent instructions. Identify the Google Cloud services, client libraries, and tests it uses, then choose the matching service guides before changing application logic.
Requires Docker running and the latest LocalCloud CLI. Install the CLI once with curl -fsSL https://local.cloud/install.sh | sh, or follow the setup guide. Skip installation when localcloud --version already works. If the installer prints a source command, run it before continuing. Setup guide
Select the project and services
For a new example directory, save this as localcloud.yaml before starting. In an existing repository, merge the relevant settings into its configuration. This example selects BigQuery, Cloud Storage, and Pub/Sub; adapt the exact service IDs to your project.
Expected result: The configuration includes every service your application needs. Firestore, GKE, and Cloud Run are opt-in.
Start or reuse the local runtime
Run from the configured directory. Start waits for readiness and can reuse the matching runtime. Export its actual mapped endpoints in the same shell or test runner that launches your SDK; do not assume the default ports are available.
setup.sh
localcloud doctor
localcloud start --local-only
eval "$(localcloud env)"
Expected result: Startup succeeds and the SDK environment points to the selected local runtime. Inspect localcloud status --verbose if startup or endpoint discovery fails.
Verify a local SDK call
Python · BigQuery. Install the client in an isolated environment, save this file, then execute a constant query through the generated local endpoint. This verifies SDK routing without creating persistent resources.
# First export endpoints: eval "$(localcloud env)"
import os
from google.auth.credentials import AnonymousCredentials
from google.cloud import bigquery
with bigquery.Client(
project=os.environ["GOOGLE_CLOUD_PROJECT"],
credentials=AnonymousCredentials(),
client_options={"api_endpoint": os.environ["BIGQUERY_EMULATOR_HOST"]},
) as client:
rows = list(client.query("SELECT 1 AS ok").result(timeout=30))
assert rows[0].ok == 1
print("ok=1")
Run the example
python assert_query.py
Expected result: The assertion passes, the command prints ok=1, and no test resources need cleanup.
Scope and references
Use LocalCloud for development, tests, and CI. If Docker or LocalCloud is unavailable, report the blocker instead of switching to real Google Cloud. Validate production endpoints, authentication, IAM, quotas, and regional behavior against Google Cloud before release.
Check Docker, start on localhost-only ports, confirm readiness, configure this shell, and open the console. If a default port is already in use, the CLI picks another one. Use the endpoint values it returns.
Precise definitions for agent environment, cloud emulator, credentialless testing, and the terms around them.
Questions
Questions developers and agents ask
How does LocalCloud compare with LocalStack and Google's own emulators?
LocalStack emulates AWS and Snowflake, not Google Cloud; teams on both clouds run both. Google ships separate emulators for Bigtable, Datastore, Firestore, Pub/Sub and Spanner through the Google Cloud CLI, and the Firebase Local Emulator Suite for Firebase products. The Google Cloud CLI has no BigQuery or Cloud Storage emulator. Single-service community emulators typically cover one Google Cloud service each, such as BigQuery or Cloud Storage; coverage and fidelity vary by project. LocalCloud runs 25 Google Cloud services, including BigQuery and Cloud Storage, in one Docker container, using Google's Firestore emulator and a Spanner emulator built from Google's open-source emulator.
Sources reviewed .
Local options for Google Cloud development compared, sources reviewed October 5, 2026
LocalCloud
Google emulators (gcloud)
LocalStack
Community emulators
Google Cloud services
25 Google Cloud services in one Docker container
Five: Bigtable, Datastore, Firestore, Pub/Sub and Spanner, each started on its own
None; LocalStack covers AWS and Snowflake
Typically one Google Cloud service each, such as BigQuery or Cloud Storage
BigQuery and Cloud Storage
Both, alongside Pub/Sub, Spanner, Firestore and the rest of the stack
No BigQuery or Cloud Storage emulator
No
Separate projects, one per service
Data across restarts
Stored in a named Docker volume that survives restarts
Pub/Sub keeps data for the emulator session; the Spanner emulator keeps data in memory only
Not applicable to Google Cloud
Varies by project
License
Proprietary Public Preview License
Google tooling; the Spanner emulator is open source (Apache-2.0)
Proprietary plans; current releases need an account and auth token since March 23, 2026
Open-source licenses that vary by project
Cost and sign-up
Free during the public preview, with no account or license key; open-source projects stay free after it
Installed with the Google Cloud CLI
Free Hobby plan for non-commercial use; paid plans for commercial use
LocalCloud is free during the public preview for individuals and organizations, including for-profit companies, for development, testing, CI, evaluation and internal pilots. You don't need an account, a payment method or a license key. Open-source projects stay free after the preview ends, and preview releases keep their terms.
Docker. The CLI runs on macOS 13+ on Apple Silicon and Linux with glibc 2.35+ (Ubuntu 22.04 equivalent), and any Docker host can run the agentcloud/localcloud image directly. The CLI gives the container 4 GB of memory by default; the quick start runs it on localhost-only ports.
Your service data stays in a Docker volume on your machine. The runtime sends pseudonymous usage and diagnostic reports; set LOCALCLOUD_TELEMETRY=false in the container's environment to stop them. Tools and jobs you connect can still send data elsewhere.
Yes. Agents get a Google Cloud target on localhost with no credentials or billing project by default, so agent-written code can't create real cloud resources by mistake. Point an agent at the agent guide, or copy the AGENTS.md template into your repository. Agent builders can write to agent@local.cloud.
Every CLI release ships SHA-256 checksums and Sigstore bundles from its GitHub Actions release workflow, and the install script checks the checksum before installing.