Solana simulation & historical data

experiment
on history,not production.

Query historical Solana state and replay history with custom logic.

Book a demoRead the docs
Use cases

Answer thewhat if questions

Tick-by-tick historical replay of onchain state, with offchain services like aggregators wired in automatically.

/data
Historical data

Pull Solana state at any slot, exactly as it was.

/regression
Regression testing

Run new code against real historical traffic before it ships.

/ab-config
A/B testing

Compare parameter sets on identical market conditions.

/liquidations
Liquidation PnL

Reconstruct liquidation outcomes and the PnL behind them.

Workflow

Rewindand replay

Same history, different logic, real results.

termina — sim
$▋
→ loading state at slot 300000000
slot 300000000 0/10
✓ 10 slots replayed with new_program.so
baseline vs. modified → diff ready
01Pick a window

Select any time range from minutes to months ago: a crash, a spike, or an edge case and replay that window.

02Adjust and test

Experiment with alternative parameters, new strategies, and competitive positioning.

03Compare results

Compare execution quality and risk metrics between the baseline and modified scenarios.

Under the hood

Change the logic.Keep everything else.

Mainnet-identical execution, counterfactual timelines, and the data to compare them.

01 · Execution engine
Mainnet conditions in, profit numbers out.

Same conditions as mainnet with custom logic.

orderingaccount statefeescompute limitscontentionoffchain services
02 · Strategy testing
Build the timeline where custom logic was live.
REPLAYWhat the live strategy earned and lost.
SIMULATIONFind opportunities that were missed and calculate the profit.
03 · Actionable data
Every change, side by side with the baseline.
→See how changes move metrics across every condition tested.
→Observe where fills improved and where transactions failed and cost money.
→Surface risks before they're real: liquidation triggers, MEV exposure, reserve pressure.

Built by engineersfrom

SpaceX
NEAR
Microsoft
Lyft
Amazon
Optiver

Replay the moment that matters.

Bring a slot range and a question. Leave with evidence.

Get in touchRead the docs
What to bring
01a slot range
02a program ID
03the binary to test
→ baseline vs. modified, same history
© Nitro Labs Inc 2026
XGitHubDocs