Market Data

Colocation Rate Benchmark

Colocation lease rates by market and workload type, $/kW/month, for a 250–500 kW requirement.

Confidence A Directly reported figure B Estimated / Derived C Regional inference / indirect estimate D Not found / insufficient data
US National Average · Enterprise/Wholesale $205 ▲ +4.6% vs 2025

$/kW/month, 250–500 kW requirement · global published range shown below

$160 · Sao Paulo (lowest)$403 · Singapore (highest)
Markets tracked 16 US, EMEA, APAC, LATAM
Directly-reported (Conf. A) 12 / 16 remainder estimated/derived

United States

$/kW/MO
MarketEnterprise/Wholesale
low – mid – high
AI
Training
AI
Inference
Media &
Broadcast
Mail/WebTrend
2017–2030
Conf.
Silicon Valley, CA
$230$250$270
$415$318$232$1702021–2030: $106–$328/kW/mo+136%2021–2026A
New York / New Jersey
$230$255$280
$423$324$237$174—C
Chicago, IL
$200$215$230
$357$273$200$1462017–2026: $140–$215/kW/mo+54%2017–2026A
Northern Virginia (Ashburn), VA
$190$213$235
$354$271$198$1452017–2030: $103–$279/kW/mo+49%2017–2026A
Northern Virginia (Ashburn), VA - 10+ MW
$160$173$185
$287$220$161$118—B
Boston, MA
$184$205$226
$340$261$191$140—A
Phoenix, AZ
$170$190$210
$315$241$177$1292025–2030: $190–$249/kW/mo+0%2025–2026A
Dallas-Fort Worth, TX
$166$185$204
$307$235$172$126—+0%2025–2026B
Atlanta, GA
$156$173$190
$287$220$161$1182017–2026: $142–$173/kW/mo+21%2017–2026B
Austin / San Antonio, TX
$148$165$182
$274$210$153$112—C
Salt Lake City, UT
$148$165$182
$274$210$153$112—C
Minneapolis, MN
$148$165$182
$274$210$153$112—C
Columbus, OH
$144$160$176
$266$203$149$109—C
Reno, NV
$144$160$176
$266$203$149$109—C
Denver, CO
$135$150$165
$249$191$140$102—B
Portland / Hillsboro, OR
$122$135$148
$224$172$126$92—B

Europe & Asia-Pacific

$/kW/MO
MarketEnterprise/Wholesale
low – mid – high
AI
Training
AI
Inference
Media &
Broadcast
Mail/WebTrend
2017–2030
Conf.
Singapore
$330$403$475
$669$512$375$2742023–2030: $300–$472/kW/mo+34%2023–2026A
Hong Kong SAR
$266$295$324
$490$375$274$201—+9%2025–2026A
Tokyo, Japan
$252$280$308
$465$356$260$1912023–2030: $200–$328/kW/mo+40%2023–2026A
Frankfurt, Germany
$235$250$265
$415$318$232$1702026–2030: $250–$316/kW/moA
London, UK
$195$213$230
$354$271$198$1452024–2030: $178–$290/kW/mo+29%2024–2026B
Amsterdam, Netherlands
$165$180$195
$299$229$167$123—B
Paris, France
$180$200$220
$332$254$186$136—C
Osaka, Japan
$180$200$220
$332$254$186$136—C
Dublin, Ireland
$166$185$204
$307$235$172$126—C
Sydney, Australia
$169$188$207
$312$239$175$128—A
Melbourne, Australia
$135$150$165
$249$191$140$102—C

Latin America

$/kW/MO
MarketEnterprise/Wholesale
low – mid – high
AI
Training
AI
Inference
Media &
Broadcast
Mail/WebTrend
2017–2030
Conf.
Queretaro, Mexico
$250$260$270
$432$330$242$177—A
Santiago, Chile
$190$200$210
$332$254$186$136—A
Bogota, Colombia
$150$190$230
$315$241$177$129—A
Sao Paulo, Brazil
$130$160$190
$266$203$149$109—A
Market Trend

Rate range by market

$/kW/MO

Markets ranked by rate, with the range behind each one. Switch workload type to compare the same markets across tenant profiles.

USInternationalLatin AmericaPublished range±10%2030 forecast

Forecast applies regional CAGRs.

Risk-adjusted price overlay

B Concept

Azraq KRI scores, each on the house 1.00–5.00 scale, weighted into a single risk number and turned back into a price. The loading is what a market would need to carry to compensate for its risk profile; the bar splits the adjusted rate into base and risk.

G Grid & power 30%W Water 15%H Hazard 15%C Community 20%M Political 20%
0.35
MarketRateGWHCMRisk Rof 5.00LoadingRisk-adjustedBase vs. risk

P index = P0 × (1 + λ × (R − 1) / 4), so a market at R 1.00 carries no loading and one at R 5.00 carries the full λ. Lambda is a decision parameter to calibrate against observed insurance, financing and mitigation costs — not a market constant. Grid, water and hazard scores come from the Azraq benchmark datasets; community and political scores are placeholders. Full framework and pricer →

Submit a deal

CONTRIBUTED COMPS

Add a rate from a lease you have signed or been quoted. Submitted deals appear on the chart above as amber diamonds and blend into a contributed average for that market.

All-in submissions are normalised to a base + metered E basis before blending, using the market's electricity rate and a 1.3 PUE. Deals are held in this browser only.