Sandbox Mode

Test your strategies against live market data without risking real money. Sandbox orders route through a simulation adapter on each venue edge that matches against the real order book.

How it works

Your order hits the SOR exactly like a live order
The SOR picks the best venue(s) based on real market data
Instead of submitting to the exchange, the edge's SandboxAdaptor simulates the fill
The fill price is realistic — it uses the actual book depth and simulates market impact
Fills, positions, and TCA all work the same as live

Tip

Sandbox fills are slightly worse than mid price (simulating taker fees + market impact). This makes backtesting more conservative than production.

Two ways to use sandbox

Option 1: Sandbox API key

Use a seq_test_* key. All orders automatically route through sandbox.

bash

export SEQ_API_KEY="seq_test_a1b2c3d4..."
sequence buy ETH-USD 50  # sandbox fill

Option 2: Per-order sandbox flag

With a live key, set sandbox=True on individual orders:

python

# Live key, but this order goes to sandbox
seq.buy("BTC-USD", 0.001, sandbox=True)

bash

sequence buy BTC-USD 0.001 --sandbox

What's real vs simulated

Aspect	Sandbox	Live
Market data (quotes, book, trades)	Real	Real
Order routing (SOR venue selection)	Real	Real
Fill prices	Simulated (against real book)	Real exchange fills
Fees	Estimated (10 bps default)	Actual venue fees
Positions & P&L	Tracked in sandbox ledger	Tracked in live ledger
Exchange API keys required	No	Yes
TCA reports	Generated	Generated

Sandbox settings

Configure sandbox behavior (requires seq_test_* key):

bash

curl https://api.sequencemkts.com/v1/sandbox/settings \
  -H "Authorization: Bearer $SEQ_TEST_KEY"

Reset sandbox state (clears positions and P&L):

bash

curl -X POST https://api.sequencemkts.com/v1/sandbox/reset \
  -H "Authorization: Bearer $SEQ_TEST_KEY"