Remote MCP · Endpoint

LLM Latency Tracker MCP server

No. LLM Latency Tracker answered an MCP handshake 1 hour ago.

UpLast checked 1 hour ago

Server URL

https://llmlatency.dev/mcp

The last check completed an MCP handshake.

Registry name
dev.llmlatency/llm-latency-tracker
Transport
Streamable HTTP
Protocol negotiated
2025-06-18
Handshake latency
86 ms
Tools exposed
2 tools
Uptime, last 30 days
100% from 17 checks
Described by its publisher as
Measured latency, time to first token and uptime for ~45 AI inference APIs, by region.
Website
https://llmlatency.dev/
Repository
https://github.com/mazamaka/llm-latency-tracker
Listing
From the official MCP registry (active).
Recent checks of LLM Latency Tracker
CheckedResultHandshake
1 hour agoHandshake completed86 ms
2 days agoHandshake completed77 ms
4 days agoHandshake completed56 ms
5 days agoHandshake completed121 ms
7 days agoHandshake completed133 ms
8 days agoHandshake completed311 ms
9 days agoHandshake completed111 ms
12 days agoHandshake completed57 ms
13 days agoHandshake completed119 ms
15 days agoHandshake completed71 ms
16 days agoHandshake completed86 ms
17 days agoHandshake completed58 ms
19 days agoHandshake completed54 ms
20 days agoHandshake completed57 ms
20 days agoHandshake completed51 ms
21 days agoHandshake completed79 ms
22 days agoHandshake completed135 ms

The last 17 checks. Full history is kept for 30 days; the all-time counter stands at 17 responding of 17.

How to read this

  • ·Handshake latency is one measurement from one probe, from our infrastructure. It is a smoke signal about the endpoint, not a benchmark of your own connection to it.
  • ·An authentication challenge counts as responding. A hosted server that refuses an anonymous prober is behaving correctly.
  • ·A tool count of “behind credentials” means the handshake worked but listing tools needed a token we do not hold.
  • ·Uptime covers the last 30 days of our own checks, at most one an hour, and says how many checks it is drawn from.

Related

All hosted MCP serversGlide connector for AI assistantsGlide for AI agents

Listing data comes from the official MCP registry. Glide is not affiliated with LLM Latency Tracker and makes no claim about it beyond what these checks measured.