Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

The best cloud and DevOps MCP servers in 2026

The best cloud and DevOps MCP servers right now are Cloudflare MCP Server, AWS MCP Server and Azure MCP Server, out of 13 we have reviewed. Each is ranked by our rating, with its price, our verdict and who it suits.

By ReviewRamble editors · Updated Nov 11, 2026 · How we review

How we ranked: by our rating out of 5, from reviews built on each product's own pricing, documentation and policies plus independent benchmarks where they exist, each checked on the date shown in the review. We take no affiliate commission.

  1. 1

    A clever answer to MCP's context problem that hands an agent the whole Cloudflare API, which is exactly why the token you give it matters more than the server.

    Best for: Teams running sites, Workers or DNS on Cloudflare who want an agent to inspect and change configuration.

    Price: Free to connect. Calls act on your Cloudflare account and count against its API rate limits; products the agent touches bill as normal.

    Read our Cloudflare MCP Server review
    4.5/5
  2. 2

    IAM condition keys that tell agent calls from human ones make this the AWS server to standardise on; the older Labs servers around it are the risk.

    Best for: Teams on AWS who want agents to explore, build and operate resources under the IAM policies and audit trail they already have.

    Price: No charge for the server or toolkit; you pay standard prices for the AWS resources your agents create or use.

    Read our AWS MCP Server review
    4.4/5
  3. 3

    Broad enough to run most of an Azure estate from a chat window, which is exactly why it should start in read-only mode with three or four namespaces.

    Best for: Teams on Azure who want an agent to query and manage resources across many services.

    Price: Free and MIT-licensed. Azure resources it touches are billed as usual.

    Read our Azure MCP Server review
    4.3/5
  4. 4

    Grafana has thought harder than most vendors about context size and permissions, and the result is a server an on-call engineer can leave attached through an incident.

    Best for: SRE and platform teams on Grafana or Grafana Cloud who want agent help during investigations.

    Price: Self-run server free (Apache-2.0). Hosted Grafana Cloud MCP counts each user as an Assistant active AI user: 3 included with Pro's $19 platform fee, then from $20 per user a month.

    Read our Grafana MCP Server review
    4.3/5
  5. 5

    Registry lookups alone justify installing it; hand it an HCP Terraform token only after deciding which of its 48 workspace tools an agent should hold.

    Best for: Teams writing Terraform with AI help, especially those on HCP Terraform or Terraform Enterprise who will configure a tool allowlist.

    Price: Free under MPL-2.0. HCP Terraform features depend on your HashiCorp plan.

    Read our Terraform MCP Server review
    4.2/5
  6. 6

    Debugging a failed deployment from the editor is where it earns its place; with 248 tools and a purchase flow, scoping the OAuth grant is not optional.

    Best for: Vercel teams who want an agent to debug, inspect and manage their deployments.

    Price: No extra charge on any Vercel plan. Hobby is free for non-commercial use; Pro is $20 per developer seat a month including $20 of usage credit.

    Read our Vercel MCP review
    4.2/5
  7. 7

    A sound default for running a dozen local MCP servers without trusting each one with your laptop, as long as you stay on a current release and accept that it is built around Docker Desktop.

    Best for: Developers and small teams who already run Docker and want many MCP servers behind one isolated, logged endpoint on their own machines.

    Price: The gateway is free and open source (MIT). Docker Desktop, which bundles it as the MCP Toolkit, is free for personal use, education, non-commercial open source and businesses under 250 employees and $10 million revenue; otherwise Pro $9, Team $15 or Business $24 per user/month billed annually. Running it on Docker Engine alone avoids the Desktop terms.

    Read our Docker MCP Gateway review
    4.0/5
  8. 8

    Handles Netlify's deploy loop competently, though an eager agent can spend a month of credits before anyone checks the bill.

    Best for: Netlify users who want their coding agent to deploy and manage sites directly.

    Price: Server is free; actions spend Netlify credits. Free plan 300 credits a month, Personal $9/month (1,000), Pro $20/month (3,000); a production deploy costs 15 credits.

    Read our Netlify MCP Server review
    3.9/5
  9. 9

    A thorough, vendor-built bridge from an agent to a ThingsBoard tenant, whose delete, attribute and firmware tools deserve a scoped account before anyone points it at production devices.

    Best for: IoT engineers and operations teams already on ThingsBoard who want to query fleets, triage alarms and inspect telemetry from an agent with a deliberately limited account.

    Price: Free (Apache-2.0). Needs a ThingsBoard instance: Cloud plans run from Free (5 devices) to Business at $749/month (1,000 devices), or self-host.

    Read our ThingsBoard MCP Server review
    3.9/5
  10. 10

    Broad and actively patched, but out of the box it hands an agent delete, exec and raw kubectl on your current context, so run it read-only against a ServiceAccount that cannot see Secrets.

    Best for: Platform engineers who want an agent to investigate clusters through kubectl, and will give it a scoped ServiceAccount and read-only mode first.

    Price: Free and open source (MIT). You pay only for the cluster it talks to.

    Read our MCP Server Kubernetes review
    3.8/5
  11. 11

    A capable operations console for teams already on Heroku, attached to a platform that has stopped adding features.

    Best for: Existing Heroku teams who want an agent to help operate apps and Heroku Postgres.

    Price: Free and Apache-2.0 licensed; no fee for the hosted endpoint. Heroku resources are billed: Eco dynos $5/month, Standard-1X $25/month.

    Read our Heroku Platform MCP Server review
    3.6/5
  12. 12

    A thin, honest wrapper around gcloud that is exactly as safe as the account behind it, which for most developers means far too powerful until you add an allowlist.

    Best for: Google Cloud engineers who want an agent to inspect and operate resources through gcloud, and will pin it to an allowlist and a narrow service account.

    Price: Free and Apache-2.0 licensed. Every command bills to your Google Cloud project as if you had typed it.

    Read our gcloud MCP Server review
    3.5/5
  13. 13

    A tidy Docker server for a throwaway machine, but it gives an agent root-equivalent control of the host with no read-only mode, and the maintainer has not answered two security reports.

    Best for: Developers and home-lab users who want an agent to run containers on a disposable or dedicated Docker host they would be happy to rebuild.

    Price: Free and open source (GPL-3.0). You pay only for the hosts running Docker.

    Read our mcp-server-docker review
    3.3/5

Sources

The main official source behind each ranked review; every review lists its full set.

More guides