Comparison
Belowdecks and Claude compared point by point
Belowdecks agents can work with Claude Code, on an Anthropic account in your company’s name, or with another AI provider. This page compares using claude.ai or Claude Code directly with what Belowdecks installs and oversees around the model.
One model, three ways to use it
Claude is Anthropic’s AI model, and your team can use it directly in two ways. Claude.ai is the assistant: each team member writes to it, hands it multi-step tasks with Cowork and can schedule recurring tasks. Claude Code and its Agent SDK are developer tools: they run an agent in the background, on a schedule, on an API call or on a GitHub event, once a developer has programmed what surrounds it. Anthropic also offers Claude Managed Agents, which hosts the agent and its runtime environment in its cloud, billed per session hour on top of tokens; what starts it and what surrounds it is still yours to program.
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. When Anthropic is the chosen provider, these agents work with Claude Code, on an Anthropic account in your company’s name; other AI providers, such as OpenAI, are possible too. With you, our team decides what starts the agents, such as an email received or a schedule, and which actions must wait for your approval, then connects the assistant on your website.
All three options can therefore build on Claude. They differ in what surrounds it: an assistant each person uses for their own tasks, a tool a developer programs, or agents our team installs and oversees to do the business’s repetitive work.
Claude.ai, Claude Code and Belowdecks, line by line
The facts about claude.ai and Claude Code come from Anthropic’s public pages, checked on October 6, 2026. Several claude.ai features depend on the plan you choose.
What you get
- Claude.ai
- An assistant for each team member
- Claude Code, Agent SDK, Managed Agents
- An agent toolkit you program
- Belowdecks
- Agents installed and monitored by a team
Who builds and maintains it
- Claude.ai
- Each user, for their own tasks
- Claude Code, Agent SDK, Managed Agents
- A developer on your team
- Belowdecks
- The Belowdecks team, from diagnostic to follow-up
What starts the work
- Claude.ai
- Your message, or a scheduled task
- Claude Code, Agent SDK, Managed Agents
- Schedule, API call, GitHub event
- Belowdecks
- An email, a schedule or a signal from your software
Your go-ahead before an important action
- Claude.ai
- Approval asked by default, adjustable
- Claude Code, Agent SDK, Managed Agents
- Permissions and hooks you program
- Belowdecks
- Set up at installation to wait for your approval
Who can approve
- Claude.ai
- The user, in their conversation
- Claude Code, Agent SDK, Managed Agents
- Whatever the developer built
- Belowdecks
- Someone on your team, also from an email
When information is missing
- Claude.ai
- Claude asks you in the conversation
- Claude Code, Agent SDK, Managed Agents
- Whatever the developer built
- Belowdecks
- The agent asks your team and waits
Assistant for your customers
- Claude.ai
- The visitor needs a Claude account
- Claude Code, Agent SDK, Managed Agents
- Through the API, built by your developer
- Belowdecks
- Assistant on your site; the customer confirms their request
Where data is hosted
- Claude.ai
- Stored in the United States
- Claude Code, Agent SDK, Managed Agents
- Your hosting or Anthropic’s cloud
- Belowdecks
- Your Belowdecks workspace; region agreed at the diagnostic
AI bill
- Claude.ai
- Included in the seat; usage-based on Enterprise
- Claude Code, Agent SDK, Managed Agents
- Claude subscription or API usage
- Belowdecks
- Billed in your name by Anthropic or another provider
Monitoring the work
- Claude.ai
- Usage analytics; audit on Enterprise
- Claude Code, Agent SDK, Managed Agents
- Logs and alerts you program
- Belowdecks
- A result for each task, a monthly report
Help and support
- Claude.ai
- Anthropic’s help center
- Claude Code, Agent SDK, Managed Agents
- Developer documentation
- Belowdecks
- The Belowdecks team
Pricing structure
- Claude.ai
- Monthly subscription per seat
- Claude Code, Agent SDK, Managed Agents
- Subscription or usage-based billing
- Belowdecks
- From $449 a month, setup quoted separately
Choosing
The situation where each one fits best
The options are not exclusive. Your team can keep claude.ai for its everyday work and hand repetitive work to Belowdecks agents.
Claude alone fits better if
Your need is mainly to write, summarize or analyze on demand, each person for their own work. Or a developer on your team wants to program agents with Claude Code and the Agent SDK, then take care of hosting and monitoring them.
Belowdecks fits better if
You want the work to get done without a person starting each task, for example when an email received or a schedule starts the agent. Important actions wait for your go-ahead, and our team installs and monitors the agents.
Common questions about Belowdecks and Claude
Does Belowdecks use Claude?
Yes, when Anthropic is the chosen AI provider. Belowdecks agents then work with Claude Code, Anthropic’s agent tool, on an Anthropic account in your company’s name. Anthropic bills you directly for what the model consumes, counted in tokens, the unit that measures the AI’s work. We do not resell tokens. The Belowdecks subscription is calculated on that same volume of tokens and pays for the software that runs the agents, along with our team’s oversight.
Belowdecks can also run its agents with other AI providers, such as OpenAI. The monthly report gives each agent’s tasks and their results, along with the tokens used and the questions handled during the month.
Our team already uses claude.ai. Should we keep it?
Your claude.ai subscription serves your employees for their own work: writing, analyzing, preparing documents. It stays useful alongside Belowdecks, whose agents work on their own, started by an email or a schedule, without a person writing to them.
The bills stay separate: your employees’ claude.ai subscription, the tokens your agents use on your company’s account with the AI provider, and the Belowdecks subscription.
Can Claude start a task when an email arrives?
In Claude Cowork, scheduled tasks run on a schedule, even when your computer is asleep or the app is closed. Anthropic documents no trigger on an incoming email. Claude Code routines start on a schedule, an API call or a GitHub event. A developer can therefore program the link between a mailbox and an API call.
In Belowdecks, an email received in the mailbox connected to your Belowdecks workspace starts the agent set up for it.
Where is the data kept in each case?
Anthropic states that data from its commercial products, including claude.ai, is stored in the United States. A Canadian region exists only through the API, using AWS Bedrock or Google Cloud Vertex.
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, Anthropic for example, on your company’s account, and may be processed outside Quebec. Your privacy impact assessment must take this into account.
Sources
The information about claude.ai and Claude Code comes from Anthropic’s public pages, checked on October 6, 2026: pricing, scheduled tasks, approvals in Cowork, sharing artifacts, where data is stored, available regions, Claude Code routines, approvals in the Agent SDK and Claude Managed Agents. Anthropic’s features and prices change often, so check them before you decide.
Claude, Claude Code and Anthropic are trademarks of Anthropic, PBC. Belowdecks can use Claude Code through each customer’s Anthropic account and is not affiliated with Anthropic.
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 Claude subscription is enough.