MCP marketplace & hosting

The trust layer for MCP.

The MCP marketplace and hosting platform built for teams that take security seriously — where every server is scored, every premium server is audited, and every call is logged.

TRUSTED BY TEAMS ATVercelLinearRaycastRetoolPosthog
github-mcp
by anthropics
94
/100
★ AUDITED2.4M calls/mo
SECURITY
23/25
RELIABILITY
19/20
DOCS
20/20
PERFORMANCE
15/15
COMMUNITY
17/20
Last audited 2 days ago
Deploy
Audit complete
0 CVEs · 0 secrets leaked
38
Audited servers
12.4M
Calls per day
147
CVEs caught
< 10m
Time to deploy
99.98%
Hosting uptime
2,400+
Teams trust Bedric
How it works

From discovery to production in three steps.

No devops burden. No guesswork on security. Your team stays productive while your CTO stays confident.

01
Browse & trust-check

Every server in the marketplace carries a public score from 0–100, built from five weighted dimensions: security, reliability, documentation, performance, and community. Premium servers are additionally audited by our team within 48 hours.

Score breakdown: Security (25) · Reliability (20) · Docs (20) · Performance (15) · Community (20)
+
02
Deploy in one click
+
03
Audit every call
+
SCORE BREAKDOWN
SECURITY92
RELIABILITY95
DOCS100
PERFORMANCE85
COMMUNITY88
TOTAL SCORE: 92/100
Trust tiers

Not all servers are equal. We make the difference visible.

Community

Open-source servers listed from the MCP ecosystem. Automated scan report available. Use with caution.

Automated CVE scan
Dependency audit
Public report
Human review
Runtime analysis
Egress mapping

Verified

Servers that have passed automated checks and basic code review. Suitable for internal tooling.

Automated CVE scan
Dependency audit
Public report
Code review passed
Runtime analysis
Egress mapping

Audited

Human-reviewed within 48 hours by our security team. Full audit report published. Safe at company scale.

Automated CVE scan
Dependency audit
Public report
Code review passed
Runtime egress analysis
48h human audit
Our solution

We solve all three with a single platform.

A curated catalog, secure managed hosting, a security-first console for IT teams, and user-friendly tools that make MCP accessible to both technical and non-technical users.

Our defensible differentiators

1
Public scoring system (0–100)

Transparent criteria with clear security signals. No other platform exposes a security score this clearly. Every server gets a public grade based on provenance, security scans, and maintenance quality.

2
Mandatory audit for premium servers

Hybrid approach: automated security checks plus 1–2 day human review. Premium = audited, period. No shortcuts, no exceptions.

3
IT-grade controls in the org console

Allowlist/blocklist management, comprehensive audit logs, security alerts on score drops or new CVEs. Give your CISO the visibility they need.

4
Hardened hosting model

Hot/cold isolation, network egress allowlist per server, runtime sandboxing. Security-first architecture from day one.

Join the Waitlist

Be among the first to access the trusted MCP marketplace. Early access includes special pricing and priority support.

No spam. Unsubscribe anytime. We'll only email you about Bedric updates.

Pricing

Three tiers, transparent, self-serve

Choose the plan that fits your team size and security requirements. All plans include our core security features and audit logging.

Free

€0

Perfect for exploration

1 member • 1 private MCP • 10K calls/month

Community catalog access
Basic security scoring
Email support
Most Popular

Team

€49/mo

For growing teams

Up to 10 members • 10 private MCPs • 500K calls/month

Everything in Free
Allowlist/blocklist controls
Audit log & alerts
Priority support

Business

€299/mo

For established teams

Up to 50 members • Unlimited private MCPs • 5M calls/month

Everything in Team
Full audit & longer retention
Custom integrations
Dedicated support

All plans include our core security scanning and transparent scoring. Overage billing at €0.30-0.50 per 10K calls beyond plan limits.

Ready to make MCP adoption safe at company scale?

Join forward-thinking CTOs and CISOs who are building the future of AI-enabled development with security and trust at the core.

FAQ

Questions? We've got answers.

Everything you need to know about MCP servers, security scoring, and getting started with Bedric.

What is an MCP server?+
MCP (Model Context Protocol) is the standard interface that lets AI assistants like Claude, Cursor, and ChatGPT connect to your tools — GitHub, Stripe, Postgres, Slack, and more. An MCP server is the adapter that bridges the AI to the API.
How is Bedrick different from Smithery or MCP.so?+
Those are directories. Bedrick is a full platform: we score, audit, host, and give your IT team visibility into every call. Our premise is that a small, trusted catalog is more valuable than a large, unsafe one.
What does 'audited' actually mean?+
Audited servers have passed both automated scanning (static analysis, dependency CVE check, egress pattern analysis) and a 1–2 day human review by our security team. The full audit report is public and linked from every server's page.
Can I publish my own MCP server?+
Yes — deploy a private server in minutes with the Free plan. Publishing to the marketplace (so others can discover and install it) is available on the Team plan and above. Paid marketplace listings with revenue share are coming in v2.
Where is my data hosted?+
All infrastructure runs on Hetzner data centers in the EU (Falkenstein, Nuremberg). Business and Enterprise plans can request dedicated isolation. We are committed to keeping multi-cloud portability open.