Cloud

The Cloud option is a fully managed model where LangChain hosts and operates all LangSmith infrastructure and services:

• Fully managed infrastructure: LangChain handles all infrastructure, updates, scaling, and maintenance.
• Deploy from GitHub: Connect your repositories and deploy with a few clicks.
• Automated CI/CD: Build process is handled automatically by the platform.
• LangSmith UI: Full access to observability, evaluation, deployment management, and Studio.

​

Get started

To deploy your first application to Cloud, follow the Cloud deployment quickstart or refer to the comprehensive setup guide.

​

Cloud architecture and scalability

LangSmith is deployed on Google Cloud Platform (GCP) and is designed to be highly scalable. Many customers run production workloads on LangSmith for LLM application observability, evaluation, and agent deployment

​

Architecture

The US-based LangSmith service is deployed in the us-central1 (Iowa) region of GCP.

​

Regional storage

The resources and services in this table are stored in the location corresponding to the URL where sign-up occurred (either the US or EU). Cloud-managed LangSmith uses Supabase for authentication/authorization and ClickHouse Cloud for data warehouse. See the Regions FAQ for more information.

​

Region-independent storage

Data listed here is stored exclusively in the US:

• Payment and billing information with Stripe and Metronome

​

GCP services

LangSmith is composed of the following services, all deployed on Google Kubernetes Engine (GKE):

• LangSmith Frontend: serves the LangSmith UI.
• LangSmith Backend: serves the LangSmith API.
• LangSmith Platform Backend: handles authentication and other high-volume tasks. (Internal service)
• LangSmith Playground: handles forwarding requests to various LLM providers for the Playground feature.
• LangSmith Queue: handles processing of asynchronous tasks. (Internal service)

LangSmith uses the following GCP storage services:

• Google Cloud Storage (GCS) for runs inputs and outputs.
• Google Cloud SQL PostgreSQL for transactional workloads.
• Google Cloud Memorystore for Redis for queuing and caching.
• Clickhouse Cloud on GCP for trace ingestion and analytics. Our services connect to Clickhouse Cloud, which is hosted in the same GCP region, via a private endpoint.

Some additional GCP services we use include:

• Google Cloud Load Balancer for routing traffic to the LangSmith services.
• Google Cloud CDN for caching static assets.
• Google Cloud Armor for security and rate limits. For more information on rate limits we enforce, please refer to this guide.

​

Allowlisting IP addresses

​

Egress from LangChain SaaS

All traffic leaving LangSmith services will be routed through a NAT gateway. All traffic will appear to originate from the following IP addresses: It may be helpful to allowlist these IP addresses if connecting to your own AzureOpenAI service or other endpoints that may be required by the Playground or Online Evaluation.

​

Ingress into LangChain SaaS

The langchain endpoints map to the following static IP addresses: You may need to allowlist these to enable traffic from your private network to LangSmith SaaS endpoints (api.smith.langchain.com, smith.langchain.com, beacon.langchain.com, eu.api.smith.langchain.com, eu.smith.langchain.com, eu.beacon.langchain.com).

​

API rate limits

LangSmith enforces rate limits on API endpoints to ensure service stability and fair usage. The following table shows the rate limits for different endpoints in both US and EU regions. Note that:

• Rate limits are expressed as count / interval where count is the number of requests allowed within the interval (in seconds). For example, 2000 / 10 means 2000 requests per 10 seconds.
• When no HTTP method is specified in the endpoint column, the rate limit applies to all HTTP methods for that endpoint.
• When a specific method is listed (e.g., POST, GET), the rate limit applies only to that method.

​

Rate limit categories

• High throughput: General high-volume endpoints for core operations like authentication, metadata, and feedback.
• Repository: Repository and prompt management operations.
• Run ingest: Individual trace/run ingestion endpoints for observability.
• Charts: Chart generation and visualization endpoints.
• Multipart ingest: Bulk run ingestion via multipart upload for high-volume tracing.
• Run query (API): API key-based run query operations with stricter limits for complex queries.
• Run query (User): User-based run query operations with higher limits for interactive use.
• Generation: AI-powered code and content generation endpoints (limited to prevent abuse).
• Commits: Prompt versioning and commit operations.
• Deletion: Session deletion and workflow trigger operations.
• Run lookup: Retrieving specific runs by UUID.
• Examples: Fetching dataset examples for few-shot prompting.
• Default (API key): Fallback rate limit for authenticated API requests not matching specific patterns.
• Default (User): Fallback rate limit for authenticated user requests not matching specific patterns.
• Public download: High-volume public download endpoints for shared resources.
• Stats: Run statistics and analytics endpoints (region-specific limits apply).
• Public (catch-all): Default rate limit for unauthenticated public access.

For more information on rate limits and other service limits, refer to the Administration overview.

---