I am evaluating FYERS API V3 WebSocket/TBT feed for high-frequency order flow analysis and would like some technical clarity regarding latency and market data throughput.
Could the FYERS team please provide measurable information regarding:
-
Average and p99 latency for Market Data WebSocket (not order execution latency).
-
Approximate tick/update throughput that a user can expect for highly active instruments such as NIFTY Futures, BANKNIFTY Futures, and ATM Options during:
-
Normal market conditions - Opening session (9:15–9:30) - Expiry days - Major event volatility
-
Does the TBT feed provide:
-
Every trade execution separately? - Every order book update separately? - Any aggregation/throttling at the broker level?
-
Is there any official benchmark available comparing:
-
Exchange timestamp vs client receive timestamp - Average WebSocket latency - Packet loss statistics - Update delivery rate during peak market activity
-
What is the maximum sustained update rate (updates/second) that FYERS infrastructure is designed to deliver for a single highly active instrument?
I found references to 50-level DOM, Protobuf-based TBT feed, and community projects mentioning 1000+ updates/sec with <10ms latency, but I would like to know whether FYERS has any official measurements or documentation for these figures.
Any detailed technical clarification would be greatly appreciated.