July 2026: sticky sessions improved market-wide
Last month we said sticky-session reliability is the metric every "top 10
proxy" list ignores. Thirty days and 368,480 probes later, the numbers moved.
The month in numbers
- 368,480 probes across 5 residential proxy providers, +77% more
measurements than the previous 30 days after we expanded probe coverage - 2 vantage points (EU + US), probes every 15 minutes, composite scores
recomputed daily - No sponsored placements. Nobody pays to move up. Methodology →
Session reliability improved
We compared the same 5 providers against the previous 30 days:
| Metric | Previous 30d | Last 30d | Change |
|---|---|---|---|
| Sticky session survival | 87.3% | 92.2% | +4.9 pp |
| Unexpected mid-session rotation | 4.9% | 3.0% | −39% |
| Overall success rate | 89.1% | 90.2% | +1.1 pp |
The spread between providers stays wide: session survival runs from 80.1%
to 98.6%, unexpected rotation from 1.0% to 6.2%. If your workflow holds
state (logins, carts), that range decides whether it survives the hour.
July's top performers
🥇 Proxyon. Best sticky sessions we measure: 98.6% survival, 1.0%
unexpected rotation. Fastest median TTFB at ~704 ms and the highest success
rate, 94.3%. Leads the live dashboard.
🥈 Maskify. 96.4% session survival, 843 ms median TTFB. Tops no single
metric and trails in none.
🥉 FleetProxy. 92.2% success rate with steady results on external
targets. A dependable pick for broad crawling.
We name leaders. Weaker results stay anonymous in the ranges above; live
scores for each provider sit on the public dashboard.
On our side
July went into polish: visual refinements across the dashboard and report
views, plus performance work. The probe-coverage expansion shows up in the
+77% above.
What's next
More providers are joining the benchmark. If you run a residential proxy
network and want in — the benchmark is free and independent —
get in touch.
Live data: proxystats.io/dashboard ·
Methodology: proxystats.io/methodology