Update the billing information for LangSmith Deployment (#1615)

Author: katmayb

src/langsmith/billing.mdx

@@ -104,12 +104,13 @@ The usage graph lets you examine how much of each usage-based pricing metric you
 
 Navigate to the usage graph under **Settings** -> **Billing and Usage** -> **Usage Graph**.
 
-There are two usage metrics that LangSmith charges for:
+There are several usage metrics that LangSmith charges for:
 
 - LangSmith Traces (Base Charge): tracks all traces that you send to LangSmith.
 - LangSmith Traces (Extended Data Retention Upgrades): tracks all traces that also have our Extended 400 Day Data Retention.
+- LangSmith Deployment Agent Runs: tracks end-to-end invocations of deployed LangGraph agents.
 
-For more details, refer to the [data retention conceptual docs](/langsmith/administration-overview#data-retention).
+For more details on traces, refer to the [data retention conceptual docs](/langsmith/administration-overview#data-retention). For more details on Agent Runs, refer to [LangSmith Deployment billing](#langsmith-deployment-billing).
 
 
 #### Invoices
@@ -179,6 +180,14 @@ For example, this is the expected configuration to keep 10% of all traces for ex
 
 If you want to keep a subset of traces for **longer than 400 days** for data collection purposes, you can create another run rule that sends some runs to a dataset of your choosing. A dataset allows you to store the trace inputs and outputs (e.g., as a key-value dataset), and will persist indefinitely, even after the trace gets deleted.
 
+### LangSmith Deployment billing
+
+In addition to traces, LangSmith charges for deployed agents via LangSmith Deployment (formerly LangGraph Platform).
+
+- **Agent Runs**: An Agent Run is one end-to-end invocation of a deployed LangGraph agent and is billed at $0.005 each. Nodes and subgraphs within a single agent execution are not charged separately. Calls to other LangGraph agents are charged separately to the deployment hosting the called agent. When using human-in-the-loop with interrupts, resuming after an interrupt creates a separate Agent Run.
+- **Deployment Uptime**: You are also charged for the time your deployment's database is live and persisting state. See the [pricing page](https://www.langchain.com/pricing) for uptime costs by deployment type (Development vs Production).
+
+For high-volume deployment usage, please [contact our sales team](https://www.langchain.com/contact-sales) to discuss custom pricing options.
 
 ### Summary
 

src/langsmith/faq.mdx

@@ -151,6 +151,6 @@ Yes! You can use the [development version of Agent Server](/langsmith/local-serv
 This will connect to the Studio frontend hosted as part of LangSmith.
 If you set an environment variable of `LANGSMITH_TRACING=false`, then no traces will be sent to LangSmith.
 
-### What does "nodes executed" mean for LangSmith usage?
+### What is an Agent Run?
 
-**Nodes Executed** is the aggregate number of nodes in a LangGraph application that are called and completed successfully during an invocation of the application. If a node in the graph is not called during execution or ends in an error state, these nodes will not be counted. If a node is called and completes successfully multiple times, each occurrence will be counted.
+An Agent Run is one end-to-end invocation of a LangGraph agent deployed via LangSmith Deployment. Nodes and subgraphs are not charged separately. Calls to other LangGraph agents (through RemoteGraph or the LangGraph SDK or the API directly) are charged separately, to the deployment that hosts the agent being called. An interrupt for human-in-the-loop creates a separate Agent Run when resuming.

src/langsmith/pricing-faq.mdx

@@ -25,6 +25,12 @@ A seat is a distinct user inside your organization. We consider the total number
 
 A trace is one complete invocation of your application chain or agent, evaluator run, or playground run. Here is an [example](https://smith.langchain.com/public/17c24270-9f74-47e7-b70c-d508afc448fa/r) of a single trace.
 
+### What is an Agent Run?
+
+An Agent Run is one end-to-end invocation of a LangGraph agent deployed via LangSmith Deployment. Nodes and subgraphs within a single agent execution are not charged separately. Calls to other LangGraph agents (through RemoteGraph or the LangGraph SDK or the API directly) are charged separately, to the deployment that hosts the agent being called. An interrupt for human-in-the-loop creates a separate Agent Run when resuming.
+
+Agent Runs are billed at $0.005 each. For high-volume usage, please [contact our sales team](https://www.langchain.com/contact-sales) to discuss custom pricing options.
+
 ### What is an ingested event?
 
 An ingested event is any distinct, trace-related data sent to LangSmith. This includes: