Measured latency to Gemini around the world
REST ticker round-trip from eight AWS regions
Gemini latency from eight AWS regions
| AWS region | Round-trip |
|---|---|
| London eu-west-1 | ~116 ms |
| N. Virginia us-east-1 | ~120 ms |
| N. California us-west-1 | ~122 ms |
| Frankfurt eu-central-1 | ~132 ms |
| Tokyo ap-northeast-1 | ~244 ms |
| Seoul ap-northeast-2 | ~283 ms |
| Hong Kong ap-east-1 | ~300 ms |
| Singapore ap-southeast-1 | ~338 ms |
Indicative REST round-trips from our multi-region measurements — for comparing regions, not an SLA. WebSocket feeds are faster in absolute terms, but the regional proportions hold.
Best server location for trading on Gemini
Host your trading server or bot in AWS London (eu-west-1): it reaches Gemini in about 116 ms, fast enough that order placement and market data stay effectively real-time.
Avoid running latency-sensitive strategies against Gemini from Singapore or other distant regions. At ~338 ms (3× the co-located figure) every price you see is stale before you can act on it, and that staleness turns into slippage on every market order.
Compare with other exchanges
Deribit server location
~10 ms
BitMEX server location
~26 ms
BEQUANT server location
~54 ms
NovaDAX server location
~58 msFrequently asked questions
Where are Gemini's servers located?
Gemini responds fastest from AWS London (eu-west-1) — about 116 ms round-trip — which places its API infrastructure in or near London.
What is the best server location for Gemini?
AWS London (eu-west-1). A server there reaches Gemini in ~116 ms, versus ~338 ms from Singapore — a 3× difference on every API call.
How much does server location change latency to Gemini?
In our measurements: ~116 ms from London versus ~338 ms from Singapore — a 3× spread. Every order and every orderbook update pays this tax in both directions.