Backblaze B2 MCP. Manage object lifecycles and audit file integrity.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Backblaze B2 MCP lets your AI client manage entire cloud storage architectures via natural conversation. You can list all buckets, adjust access rules, find failed multipart uploads, or deeply inspect file metadata and versions without leaving your IDE.
What your AI agents can do
Authorize account
Gets the specific API endpoint needed to connect your agent to your unique B2 billing account region.
Create bucket
Provisions a brand new, empty logical storage bucket partition within Backblaze.
Delete bucket
Irreversibly deletes an entirely empty B2 Storage Bucket after confirming no files remain inside.
List all existing global storage buckets and verify the necessary API endpoints for your account.
Create, update access rules on, or permanently delete entire storage buckets.
Retrieve granular details for any object, including its size, content type, and unique SHA1 checksum.
Soft-delete files by marking them hidden or hard-delete specific historical versions to free up space.
Scan for stalled, incomplete large file chunks left over from failed multipart uploads.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
Backblaze B2 with 10 Tools
These tools give your agent direct access to every function needed to manage backblaze b2 buckets, files, and object lifecycles.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using Backblaze B2 on Vinkius019d7557authorize account
Gets the specific API endpoint needed to connect your agent to your unique B2 billing account region.
019d7557create bucket
Provisions a brand new, empty logical storage bucket partition within Backblaze.
019d7557delete bucket
Irreversibly deletes an entirely empty B2 Storage Bucket after confirming no files remain inside.
019d7557delete file version
Permanently removes a specific, absolute historical chunk of data from the storage arrays.
019d7557get file info
Retrieves detailed metadata about a file, including its size, content type, and calculated SHA1 checksum.
019d7557hide file
Marks an active B2 file as hidden without deleting the binary payload, useful for safe soft deletion logic.
019d7557list buckets
Retrieves a complete list of every storage bucket currently existing globally on your account.
019d7557list file names
Paginates the primary surface-level object metadata, showing visible files within a specified bucket.
019d7557list unfinished large files
Scans B2 nodes to locate and report on stalled multipart upload chunk aggregates that failed completion.
019d7557update bucket
Changes the global access control list (ACL) privacy settings for a defined storage bucket.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Backblaze B2, then connect any of our 4,800+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,800+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Backblaze B2. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
The headache of cloud console navigation
Today, managing storage means opening the web console, navigating to the correct bucket, clicking into file lists, and sometimes running separate scripts in your terminal just to check if a large upload finished or if an old version is still taking up space. It's slow; it requires context switching between three different interfaces.
With this MCP, you talk to your agent once. You ask it to 'audit the storage.' The agent handles the multi-step process—checking for failed chunks using `list_unfinished_large_files`, listing all containers with `list_buckets`, and reporting back a unified status summary. It just works.
Controlling Object Lifecycles with `delete_file_version`
Before this MCP, removing old data was risky. You might delete the main file but miss a historical version, or worse, accidentally delete an active object because you couldn't verify its unique ID. The process felt like playing Russian roulette with your data.
Now, you ask the agent to `delete_file_version` and specify the exact chunk. It handles the irreversible removal of that specific piece of data from disk arrays, letting you prune history precisely without touching anything else.
What you can do with this MCP connector
Managing object storage often means jumping between a clunky web console, running shell commands, and manually checking logs for broken transfers. This MCP lets you run those complex tasks directly through your agent. Need to audit which buckets exist, check if an old version of a file is still taking up space, or create a new private domain? Just ask.
It handles the orchestration of secure object storage so you don't have to live in the cloud provider's UI. With this MCP connected via Vinkius, your agent can treat Backblaze B2 like another service endpoint, giving you full control over buckets, file versions, and access policies right inside your editor.
019d7557-f367-7302-9200-58f220cb6eb8 
How Backblaze B2 MCP Works
- 1 First, subscribe to this MCP and enter your Backblaze B2 Application Key ID and Key.
- 2 Next, run a command through your AI client asking it to perform a storage task (e.g., 'List all buckets' or 'Create a new private bucket').
- 3 The agent executes the necessary calls against B2, returning structured data about the status of the bucket, file, or upload stream.
The bottom line is you talk to your agent; it talks to Backblaze B2 for you.
Who Is Backblaze B2 MCP For?
This MCP is critical for Ops Engineers who spend too much time in the cloud console, or Backend Developers needing to provision secure storage domains on the fly. If your job involves managing data pipelines and auditing object integrity, you need this.
Needs to rapidly verify bucket states, clean up abandoned file versions, or debug failed multipart upload streams after a deployment failure.
Requires the ability to create public and private storage domains directly from their development environment without context switching. They use it to prototype secure asset handling.
Needs to audit object footprints securely, perform granular file integrity checks using SHA1, or adjust global ACL privacy settings for compliance reasons.
What Changes When You Connect
- Don't waste time in the console checking for failed uploads. Use
list_unfinished_large_filesto scan your entire account for stalled multipart upload chunks, immediately pointing you toward pipeline failures. - Need to provision a secure area fast? Simply ask your agent to use
create_bucketand specify if it needs to be allPublic or allPrivate; the setup is instant. - Tired of worrying about old data versions cluttering up costs? You can run
delete_file_versionto vaporize specific historical chunks, giving you granular control over your storage footprint. - File auditing used to mean running complex CLI commands. Now,
get_file_infolets your agent pull the SHA1 checksum and content type for any file in plain language. - When a client needs temporary access to a document but it shouldn't be permanently listed, use
hide_file. It enforces safe soft deletion logic without losing the data itself.
Real-World Use Cases
Debugging a corrupted asset pipeline
A DevOps engineer suspects that some large video files are failing to complete upload. Instead of manually checking logs, they ask their agent to run list_unfinished_large_files. The agent finds the broken chunks and reports exactly which bucket needs attention.
Setting up a secure staging environment
A Backend Developer needs a dedicated area for testing public assets. They instruct their agent to use create_bucket with specific ACL rules, ensuring the new container is provisioned correctly and immediately.
Auditing old data compliance
A System Administrator must ensure that a sensitive bucket only contains files marked as private. They run list_buckets to verify the structure, then use update_bucket to enforce strict ACLs across all containers.
Recovering from accidental public exposure
A developer accidentally uploaded a temporary file that was visible globally. They ask their agent to first run list_file_names to confirm the object name, and then use hide_file to immediately remove its visibility without deleting the underlying data.
The Tradeoffs
Deleting a bucket with active files
Trying to run a delete command on a bucket that still contains even one file. The operation will fail because B2 requires cleanup first, leaving the user stuck.
→
First, list all visible names using list_file_names to confirm content exists. Then, if necessary, use hide_file on the assets you want to keep before running delete_bucket.
Assuming all files are always accessible
Relying solely on standard file listings when debugging a pipeline failure. You might miss shadowed or versioned chunks that still exist.
→
Always pair list_file_names with a call to get_file_info. This ensures you retrieve the full SHA1 checksum and content type, giving you total data integrity confirmation.
When It Fits, When It Doesn't
Use this MCP if your workflow involves frequent object lifecycle management: creating domains, auditing specific files, or debugging failed transfers. If you need to know what's in a bucket right now, list_file_names is your go-to tool. However, don't use it if you only need basic file sharing; standard cloud tools are fine for that. You must use this MCP when the task involves deep auditing or structural changes: check get_file_info for checksums, and use list_unfinished_large_files any time a large upload fails.
Common Questions About Backblaze B2 MCP
How do I list every bucket using the `list_buckets` tool? +
You simply ask your agent to run list_buckets. It returns a structured report of every bucket ID on your account, letting you see everything at once.
What is SHA1 and why do I need the `get_file_info` tool? +
SHA1 is a unique digital fingerprint (checksum) for a file. You use get_file_info to generate this checksum, which lets you verify that an object hasn't been corrupted or tampered with since it was uploaded.
Can I delete a bucket using the `delete_bucket` tool if files are inside? +
No. The delete_bucket tool will fail if any standard files persist inside. You must first use tools like hide_file or ensure all data is cleared before attempting deletion.
What does the `list_unfinished_large_files` tool check for? +
This tool scans B2 nodes specifically looking for stalled Multipart Upload chunk aggregates. It helps you debug failed large file transfers that didn't finish uploading correctly.
How do I find my specific regional data ingress point using the `authorize_account` tool? +
It uncovers your dynamically assigned, region-specific API URL. This is critical because it guarantees you're hitting the correct global session routing endpoint for accurate performance and billing.
What does the `hide_file` tool do, and how is that different from deleting a file? +
It doesn't delete the binary payload; it injects a null-marker that makes the file invisible to standard listing tools. You can safely mark an active B2 file as hidden while keeping the data intact for later recovery.
How do I change if my bucket is public or private using `update_bucket`? +
You mutate the global Access Control List (ACL) privacy settings for an entire storage bucket. This lets you switch between allPrivate and allPublic statuses without having to recreate the partition.
When should I use `delete_file_version` instead of just listing files? +
Use this when you need to permanently remove a specific version of data, avoiding long-term retention costs. It vaporizes an absolute chunk data footprint from disk arrays.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.