PredictIt rate limits
As of October 26, 2026, PredictIt warm book latency is 121 ms at the median (p50, 24h window), with 232 ms at p90 and 950 ms at p99. PredictIt ranks #2 of 6 chains measured on this benchmark.
PredictIt enforces a published rate limit of 5 requests per minute (one request per 12 seconds) per IP. The harness probes at 15 seconds to stay within budget with a safety margin. This makes PredictIt the tightest rate-limited venue in the cohort by a wide margin. The daily ramp is disabled for PredictIt: running any ramp tier above 4 requests per minute would immediately exceed their documented budget. The benchmark records latency at the steady-state probe rate only.
How PredictIt compares
Live p50 over the last 24 hours across every chain on this benchmark, ranked lowest first.
Methodology, charts and the full ledger live on the Prediction market API rate limits, tested with a daily ramp page. Raw data: JSON endpoint.