# What llmlatency.dev runs on, and where to run your own AI agent

Updated November 07, 2026.

## What the measurements run on

- **Vultr** (Probe servers · US, Tokyo, São Paulo): Three of the four probe nodes behind every number on this site are Vultr cloud instances. Every five minutes each one opens a real connection to each provider’s API and records the answer. Chicago, Tokyo and São Paulo regions; the cheapest instance size is enough for a probe.
- **Hetzner** (Probe server · Europe, database, site build): The Europe probe, the measurement database and the build that produces these pages run on one Hetzner cloud server in Nuremberg. Hetzner closed its referral programme in 2026, so this link earns nothing. It is here because the site runs on it.
- **Cloudflare** (Hosting · CDN · edge): Cloudflare Pages serves the static site; a Cloudflare Worker serves the redirects and counts partner clicks without cookies. No referral programme for Pages or Workers.

## Where should I host an AI agent, monitor an API or rent a GPU?

- **Hyperping** (Uptime monitoring for your endpoints): Watch your own API, webhook or status page the way this site watches providers: checks from several regions, alerts and a public status page. This site does not use Hyperping; its probes are our own code on our own servers.
- **Railway** (Hosting for an AI agent without a VPS): Run OpenClaw, Hermes Agent, n8n or any container on a managed platform instead of patching a VPS. A new account starts with $20 in credits. This site does not run on Railway; the probes are plain Hetzner and Vultr VPSes.
- **Vast.ai** (GPUs by the hour for open-weights models): Rent a GPU on a marketplace when you want to run Llama, Qwen or DeepSeek weights yourself rather than call a hosted API. This site measures hosted APIs only and does not use Vast.ai.
- **Novita AI** (A measured provider with a partner programme): Novita is one of the inference APIs measured here, by the same probes and the same code as every other. The only measured provider whose programme we have joined. The link never appears in a ranking.

## What infrastructure does llmlatency.dev run on?

Three probe nodes on Vultr (Chicago, Tokyo, São Paulo), one Hetzner server in Nuremberg for the Europe probe, the database and the site build, and Cloudflare Pages and Workers for hosting.

## Are the partner links on llmlatency.dev tied to the rankings?

No. Partner links are marked, routed through /go/<name>, and never placed inside a ranking, table or comparison. Commissions have no input into any measured figure.

## What is the best VPS for an AI agent?

The measured answer is that the region matters more than the vendor: the per-region time to first byte for the LLM APIs built into OpenClaw, Hermes Agent and n8n is on llmlatency.dev/self-hosted-agents. This site's own probes run on Vultr (Chicago, Tokyo, São Paulo) and Hetzner (Nuremberg).

## Where can I host OpenClaw or n8n without running a VPS?

A managed container platform such as Railway runs the agent without server upkeep; a new account starts with $20 in credits. This site itself does not run on Railway.

## Why does the site recommend products it does not use?

Each card says whether the site uses the product. Hyperping, Railway and Vast.ai are listed because they solve problems the site's readers have; the cards state plainly that the probes run elsewhere.

Partner links are HTML-only and listed at https://llmlatency.dev/methodology#affiliate. Source: https://llmlatency.dev/stack