Proxy-to-account ratio for Instagram in 2026 — how many accounts per port without correlation

Why this is the question that decides farm economics

Proxies are ~95 % of a farm's monthly bill (see the scaling cost model). So the ratio you can safely run — accounts per proxy port — is the single biggest lever on profitability. Push it too high and survival collapses; keep it too conservative and you're burning money on idle ports. The answer isn't one number — it depends on what the accounts are doing.

What IG actually correlates

IG's anti-fraud model doesn't ban "accounts sharing an IP" outright — millions of legitimate users share carrier-grade NAT IPs. What it punishes is correlated behaviour from a shared network: multiple accounts on one IP all taking write actions in the same window, with similar fingerprints and rhythms. The IP is the thread that lets it cluster them. So:

• Passive activity (scrolling, story-views, watching Reels) is low-correlation. You can stack more accounts per port.
• Write activity (follow, DM, comment, post) is high-correlation. This is where you need isolation.

The ratio table by task type

Account task	Safe ratio	Why
Mass DM / follow (write-heavy)	1:1	Highest-watched actions; correlation = checkpoint
Warm-up (mixed, low write)	2:1 – 3:1	Mostly passive; stagger the few writes
Story-view / engagement only	3:1 – 5:1	Passive, low correlation footprint
Parsing / scraping (no writes)	5:1 – 10:1	Read-only barely touches the trust model

Rule of thumb: the more write actions per account-day, the closer to 1:1 you must go. When in doubt for active DM accounts, 1:1 is never wrong — it's the cost of doing serious volume.

ASN diversity beats raw IP count

This is the most-misunderstood part of proxy buying. A provider advertising "200 IPs/day" sounds great until you discover 170 of them cycle within the same /23 ASN. To IG's correlation model, those 170 IPs are effectively one neighbourhood. What matters is unique ASN per day, not raw rotation count. We covered this in the mobile-proxy benchmarks — Provider B did 200 IPs/day but only ~30 distinct ASNs.

For a 500-account farm, don't put all active accounts behind one provider. Spread across 2-3 providers so a single ASN going dirty (or a provider getting flagged) doesn't take down your whole cohort.

Rotation: schedule it, don't randomize blindly

• Match rotation to account rhythm. Rotating IP mid-DM-session looks like session hijacking. Rotate between sessions, not during.
• Sticky for the session, fresh per day. The ideal pattern: one stable IP for the duration of a session, a fresh ASN for the next day's session.
• Never rotate during warm-up. Continuity is part of the trust you're building. See the warm-up protocol.

A quick self-check before you scale a ratio

1. Run a 20-account canary cohort at the new ratio for 14 days.
2. Tag it and watch day-7/14 survival vs your 1:1 baseline.
3. If survival drops more than ~5 points, the ratio is too aggressive for that task type — back off.
4. Use our Proxy Checker to confirm every port is actually live, anonymous, and not leaking your real IP before you assign accounts to it.

Where this lives in the course

Module 8 of IAM Mastery covers proxy architecture: the per-task ratio map, ASN-diversity sourcing, and the rotation schedule wired into IAM's proxy manager. We also resell vetted private mobile proxies at < 30 % of typical margin — see /proxies.