Wasabi MCP Beta is Live: Your AI Agents Now Have Direct Access to Cloud Storage

Getting an AI agent to work with cloud storage has always meant writing the integration yourself. Scripts, hardcoded endpoints, manual API calls: hours of setup before a single operation could run.

Wasabi MCP changes that, and your economics. Available now in beta, Wasabi MCP gives any AI agent direct, standardized access to your entire Wasabi environment. One connection, over 140 tools, no custom code required, and flat-rate storage pricing that gets more valuable as your agent workloads scale.

What is MCP and why does it matter for cloud storage?

Model Context Protocol (MCP) is an open standard that defines how AI agents connect to external systems, including backend storage, through natural language. For Wasabi, that means one secure MCP connection gives any AI agent or LLM immediate access to Wasabi storage through a standardized set of tools, using nothing but natural language. What previously required custom scripts, hardcoded endpoints, and manual API calls for every new workflow is now a single natural language prompt. The agent builds a plan, executes each step in sequence, confirms before making any permanent changes, and reports back, without the developer touching a console or writing a line of integration code. 

What does Wasabi MCP give AI agents access to?

Wasabi MCP provides over 140 tools across three service families:

• S3 Operations: Bucket management, object operations, lifecycle policies, access controls

• Identity Access Management (IAM) : Users, roles, access policies, account management

• Wasabi Account Control Manager (WACM): Sub-accounts, members, billing, usage reporting

Wasabi MCP gives agents operational control across storage, identity management,  and account governance as a unified layer. Agents connect once via OAuth; credentials stay encrypted on the server and the AI client never sees them. No S3 API knowledge required.

Here are examples of what that looks like in practice:

• Ask the agent to generate a download link for all objects with an .mp4 extension from a bucket; it hands back a short-lived pre-signed URL for each one without the developer touching the console or running a single CLI command.

• Ask it to onboard a read-only contractor; it creates the read-only policy, a new group, adds the user to the group, attaches the policy, and  commits the changes. Previously, that sequence required scripting each step individually.

• Query the full account hierarchy in plain language: total channel accounts, sub-account counts, top accounts by purchased storage, individual member roles, and MFA status. All from a structure that would otherwise require significant console navigation or custom reporting.

How do you deploy Wasabi MCP?

Wasabi MCP is available in two models:

Wasabi Hosted MCP: Wasabi deploys and manages the MCP server on your behalf. Connect your AI agent to a pre-configured Wasabi endpoint authenticated via OAuth 2.0.  There is no infrastructure to set up, and no server to maintain, just a secure, direct connection to your Wasabi storage account. OR

Self-Hosted MCP: Deploy and operate the MCP server in your own environment. Download the executable here & follow the setup instructions. You maintain full control over your configuration and infrastructure; ideal for teams with compliance requirements or specific deployment needs.

Both models provide access to all tools and capabilities and are available now in beta with a Wasabi storage account.

You can have your first agent connected and running in four steps:

1. Sign in to your Wasabi account (or start a free trial) and get your S3 key set.

2. Connect your LLM or AI agent to your Wasabi account securely via OAuth 2.0.

3. Prompt your agent to perform the desired action against your Wasabi account.

4. Wasabi MCP orchestrates your prompt and executes appropriate actions.

What does Wasabi MCP cost?

Wasabi MCP is beta feature available at no additional charge. You pay only for your standard Wasabi cloud storage.

AI agents read and retrieve data continuously. On platforms that charge API request and egress fees, every agent interaction generates a cost that compounds unpredictably at scale. Wasabi charges no egress fees and no API fees, so agent workflows run at a flat, predictable cost regardless of how frequently your agents access your data. As agent usage grows, the cost structure does not change.

How do you get started?

Explore the quick start guide to connect your first agent. Pick a use case, connect, and see what your storage can do when agents are running it. For a deeper technical reference, the Wasabi MCP documentation has you covered.

Not a current Wasabi account holder? Start your 30-day free trial or buy now.