Product overview
From a market observation to evidence you can revisit.
Test a crypto setup against recorded history, challenge the result and decide what deserves another look. Research, replay and live order flow work together.
Test a setup
Describe a question, review its conditions and scope, then read counts and forward outcomes beside the editable definition.
Explore →Challenge the result
Compare with a same-scope reference when available. Check missing outcomes, market concentration and examples that disagree.
Explore →Replay a move you missed
We record every Binance USDT-M perpetual symbol. Revisit available trades and book updates in your browser without running your own recorder.
Explore →Save and revisit
Keep a study in the Workbench. Confirm monitoring when a supported definition is ready for forward observation.
Explore →Research from an assistant or code
Use the same engine through MCP or REST, with exact query documents and machine-readable results.
Explore →Watch live order flow
Read charts, DOM, tape, depth and liquidation layers across Binance USDT-M and Hyperliquid perpetuals.
Explore →Explore market events
Browse named historical events and use their context to form a question worth investigating.
Explore →Learn through real examples
Work through replay lessons and practice reading a market before the recorded outcome is revealed.
Explore →Live views and recorded history have different coverage
Live coverage includes Binance USDT-M and Hyperliquid perpetuals; current searchable research covers Binance USDT-M. Available dates vary by market and measurement. A longer candle-based study may reach beyond recorded order-book history.
Pro tick replay covers 90 days. Research reaches the available recorded replay history. Public demos and named archived events have their own access rules. Check data coverage and compare plans.
Read each liquidation layer for what it measures
The modeled liquidation field estimates pressure. The Hyperliquid position layer uses covered real positions and exchange-reported liquidation prices. Realized liquidations are feed observations. None is a complete map of every leveraged position.
Built for investigation
A study reports historical evidence, not a trading verdict. Outcomes can overlap, references are descriptive and missing windows remain visible. A negative or inconclusive result is a useful reason to refine a question or stop pursuing it.
Prefer your own setup? Explore the open-source terminal and gateway, including local replay tools. The hosted research corpus and services are separate.