State of MCP, 2026-07-02 census

Most MCP servers are not dead. We checked all 14,559.

A census of the entire official MCP registry, checked with the same zero-key probes anyone can run themselves. About 79.6% are healthy, roughly 1 in 5 is dead or broken, and only about 1 in 25 is fully dead. A verdict here is a measurement, not a rating.

Census date
2026-07-02
Servers checked
14,559
License
CC BY 4.0
Probe engine
akashi v0.3.0

The verdict distribution

11,582

healthy (79.6%)

2,212

degraded (15.2%)

596

dead (4.1%)

169

unknown (1.2%)

Dead-or-broken (degraded plus dead) is 2,808 servers, roughly 1 in 5. Fully dead is 596, roughly 1 in 25. Both counts are over the full 14,559, the more conservative and complete denominator.

Honest denominator note: 169 servers declared only entrypoints we cannot probe without a key, so they are reported as unknown rather than folded into a rate. Excluding them narrows the base to 14,390 servers; over that smaller base healthy rises to 80.5% and dead-or-broken to 19.5%. We lead with the full 14,559 above because it is the more conservative, complete denominator.

Failure concentrates differently per layer

A single verdict hides where things actually break. Every server can fail at up to four independent layers, and they do not fail at the same rate.

Packages: the most reliable layer

Of 9,029 package entrypoints we could probe across npm, PyPI, and OCI, only 86 are broken, about 1%. A further 971 are OCI or private images we cannot probe without a key; they are reported separately, not counted as broken.

Repositories: the rot layer

Of 12,308 declared GitHub repositories, 1,621 return HTTP 404 or are gone and 81 are archived. Combined, 1,702 repositories have rotted, 13.8% of repository-bearing servers. A further 1,675 repository checks came back an ambiguous error, possibly transient or rate-limit related. We exclude those from the rotted figure instead of inflating it.

Hosted remotes: the fragile layer

Of 7,158 declared remote endpoints, 975 are down, 13.6%: unreachable, a server error, or not found. Remote-bearing servers (6,809 of them) are the sickest cohort: 25.9% dead-or-broken and 7.5% fully dead, versus 4.1% fully dead overall. 1,762 endpoints are auth-gated: alive, but not verifiable without a key. 50 answer HTTP 200 but are not real MCP servers, caught by an impostor guard before they ever count as healthy.

server.json validity

305 servers, 2.1% of the 14,559 that declare one, publish a server.json that fails validation against its own declared JSON Schema.

Conformance, not just uptime

3,914 remotes complete a full MCP initialize handshake. Of those, 3,719 go on to resolve tools/list over the official go-sdk client, the strongest keyless proof a server actually works.

Keyless and reproducible

Keyless

Every probe touches only public endpoints: the registry, the public GitHub API, npm, PyPI, anonymous Docker Hub, and a capability-only MCP initialize / tools-list handshake. Akashi authenticates to no probed server and runs none of its tools. A GitHub token, if present, only raises the public API rate limit.

Reproducible

Same registry, same probe set, same documented criteria, akashi v0.3.0. Any row is re-checkable with akashi check <server>, any time.

Data access

Open data, CC BY 4.0, dated 2026-07-02.

Check it yourself

$ akashi check io.github.owner/name

Install Akashi - single Go binary, zero keys, MIT licensed.