Comparison
Belowdecks and LangGraph compared point by point
LangGraph gives your developers a framework to program agents. Belowdecks installs AI agents in your business, and our team sets them up and oversees them. This page compares the two on the points that matter to a small business.
Two ways to hand work to AI
LangGraph is a programming framework released under the MIT license by LangChain. It is used to build stateful agents with durable execution: an agent can stop, keep its state and resume later. It is aimed at developers, in Python or TypeScript. You deploy the whole thing on your infrastructure or on LangSmith, LangChain’s commercial platform. On LangSmith, an agent programmed with LangGraph starts through the API or on a schedule, and other triggers have to be developed. LangChain also sells LangSmith Fleet, where you create agents without code, started by Gmail, Slack, Microsoft Teams or a schedule, with optional approval before each tool. LangChain also publishes Agent Chat UI, an MIT-licensed chat interface that your developers deploy and adapt for your team or your customers.
Belowdecks installs AI agents in your business that take on part of the work, and our team sets them up and oversees them every month. With you, it decides what starts each agent and which actions must wait for approval by someone on your team. An agent does a piece of work from start to finish, such as answering a quote request received by email, and asks your team when information is missing.
LangGraph therefore suits a team of developers who want fine control over their agents. Belowdecks suits a business that wants to hand work to agents without writing code.
LangGraph and Belowdecks, line by line
The facts about LangGraph come from its public pages, checked on October 6, 2026.
What you get
- LangGraph
- A programming framework for your agents
- Belowdecks
- Agents installed and monitored by a team
Who builds and maintains it
- LangGraph
- Your developers, in Python or TypeScript
- Belowdecks
- The Belowdecks team, from diagnostic to follow-up
What starts the work
- LangGraph
- API, cron on LangSmith; Gmail, Slack, Teams with Fleet
- Belowdecks
- An email, a schedule or a signal from your software
Your go-ahead before an important action
- LangGraph
- interrupt() in the graph, which you program
- Belowdecks
- Set up at installation to wait for your approval
When information is missing
- LangGraph
- Pause and resume, programmed in the graph
- Belowdecks
- The agent asks your team and waits
Interface for your team
- LangGraph
- Agent Chat UI, MIT-licensed, to deploy
- Belowdecks
- The agents’ questions and approval requests
Assistant for your customers
- LangGraph
- You build it, or adapt Agent Chat UI
- Belowdecks
- Assistant on your site; the customer confirms their request
Where data is hosted
- LangGraph
- Wherever you deploy, or LangSmith
- Belowdecks
- Your Belowdecks workspace; region agreed at the diagnostic
AI bill
- LangGraph
- The provider your code calls
- Belowdecks
- Billed in your name by the AI provider
Pricing structure
- LangGraph
- Free (MIT); LangSmith per seat and usage
- Belowdecks
- From $449 a month, setup quoted separately
Choosing
The situation where each one fits best
Both options meet real needs. The right choice depends mostly on the time and skills your team has.
LangGraph fits better if
Your team includes Python or TypeScript developers who want fine control over how the agents behave, to integrate them into your own software and run them in production. You are ready to develop the triggers and adapt the interface.
Belowdecks fits better if
You have no developer to dedicate to the agents. You want important actions to wait for your go-ahead, and our team to install, update and monitor the agents every month.
Common questions about Belowdecks and LangGraph
How does LangGraph handle human approval?
The interrupt() function pauses the graph before an action, such as sending an email. The program then resumes with Command(resume=…), possibly with arguments changed by the person. A checkpointer is needed to keep the state during the pause. Your developers decide where to place these pauses and build or adapt the screen where a person answers.
With Belowdecks, we set up the agents at installation to ask someone on your team for their go-ahead before an important action. Depending on the setup, the request comes up in a conversation, during the agent’s work or before it starts, and that person can answer from an email. Before it goes live, each agent is tried on real cases, in a disposable workspace and instructed to send and change nothing, and you read what it would have done.
Can our developers build the equivalent of Belowdecks with LangGraph?
Yes, if they have the time. LangGraph provides the orchestration of the agents, and LangSmith adds schedules, tracing and alerts; Fleet offers a few triggers and approvals without code. The rest has to be developed and then maintained: starting the agents from your own tools, the assistant for your customers and the limit that keeps each agent in its folder.
Belowdecks provides these pieces, and our team sets them up with you at installation.
Where is the data kept in each case?
With LangGraph, data stays wherever your team deploys the application, or on LangSmith if you use that platform.
Your data is kept in your Belowdecks workspace, hosted in a region agreed at the diagnostic, which may be outside Quebec. The text an agent reads goes through the AI provider, such as Anthropic or OpenAI, on your company’s account, and that provider may process it outside Quebec. Your privacy impact assessment must take this into account.
How do the costs compare?
LangGraph is free under the MIT license. LangSmith offers a free Developer plan, a Plus plan billed per seat and by usage, for traces and deployments, and an Enterprise plan on quote. Add the model calls and the development and maintenance time.
With Belowdecks, plans start at $449 a month with Crew, which includes 15 million tokens, and the setup is quoted separately. It includes the agents our team designs for you, and the number of agents does not change the price. The AI provider bills you directly for what the model consumes. The Pricing page details the three plans.
Sources
The information about LangGraph comes from its repository, its documentation and LangSmith’s pricing, checked on October 6, 2026: GitHub repository, interrupts, cron jobs, Agent Chat UI, LangSmith Fleet triggers and pricing. Features and prices change often, so check them before you decide.
LangGraph and LangSmith are LangChain products. Belowdecks is not affiliated with LangChain.
First call
Find out whether some of your work suits agents
A free 20-minute call to describe the work that takes up your time. We will tell you whether Belowdecks fits, or whether your developers can build it with LangGraph.