Know how many devices share an IP
Go beyond “residential” or “mobile.” See whether an IP is used by 1 device, 10, or 10,000+ and whether sharing is concurrent or rotating over time.
Request Early AccessDynamic IP: 1 device at a time, but ISP rotates assignments over the month. Safe for short-window targeting.
6 device count buckets
Each IP is classified into a logarithmic bucket based on observed devices. Buckets represent order-of-magnitude ranges — more statistically robust than false-precise exact counts.
Interpret device sharing over time
Compare daily, weekly, and monthly windows to distinguish always-shared infrastructure from rotating dynamic IPs.
Daily
Security blast radius, real-time blocking decisions
Weekly
Campaign-duration measurement, short-term targeting
Monthly
List validation, attribution, long-term patterns
How to read across windows
If an IP shows 1 device daily but 10 monthly, it's a dynamic IP with ISP rotation — one device at a time, but different ones over the month. If an IP shows 100+ devices daily, it's concurrently shared (CGNAT, VPN, university).
Validated across major global carriers
Device Count accuracy has been verified on known CGNAT networks across multiple countries and carrier architectures.
Spark New Zealand
New Zealand117 heavy-sharing networks — #1 globally by CGNAT network count
Bell Canada
Canada37 heavy-sharing networks — Includes hyper-scale (100K daily) pools
NTT Docomo
Japan49 heavy-sharing networks — Major mobile carrier CGNAT
UNINET / Telmex
Mexico27 heavy-sharing networks — Many smaller CGNAT pools across Latin America
Clear, granular and consistent data
Device Count records include network-level identifiers and three time-windowed device count buckets. Available as database downloads in CSV, JSON, and Parquet formats.
Sample Record
Data Fields
Weekly data refresh · Schema is exploratory (field names not final).
network
Network range detected (e.g. 1.2.3.0/24)
asn
Autonomous System Number (e.g. AS7922)
shared
Shortest window with significant sharing (>10 devices): daily, weekly, or monthly
devices_daily
Observed unique devices per day (logarithmic bucket)
devices_weekly
Observed devices over 7 days (logarithmic bucket)
devices_monthly
Observed devices over 30 days (logarithmic bucket)
What you can build with Device Count
Physical-network scale context for IP intelligence — without cookies, fingerprinting, or invasive tracking.
Security & Fraud
Calculate blast radius before blocking. Auto-block single-device IPs; flag carrier-scale sharing for review.
AdTech & Measurement
Filter hyper-performing IPs from attribution. Detect when 1,000 "conversions" came from one CGNAT IP.
Data Quality
Validate third-party IP lists for CGNAT and MDU contamination. Diagnose why lists underperform 20x.
Network Operations
Build CGNAT-aware rate limiting. Adjust thresholds proportionally to observed device counts.
Get early access to Device Count
We're selecting a small group of companies to test Device Count during alpha. Get direct access to the IPinfo data team and help shape the product.
7.7M networks analyzed
1.3M networks with daily sharing signals, covering major carriers and ISPs worldwide.
Weekly data refresh
Device count buckets are updated weekly during alpha, with faster cadence planned.
Database downloads
Available in CSV, JSON, and Parquet formats for easy integration into your pipeline.
Direct access to the data team
Work directly with IPinfo's data team to shape the product and provide feedback.