Evidence
The methodology behind every number on this site
We’d rather you didn’t take our word for any of this. Here’s exactly where each figure comes from, and what it doesn’t tell you.
Enso Shield Independent
Twenty aggregators, six chains, a new test cycle every fifteen minutes, quotes requested in parallel with identical inputs. Every quote is re-simulated against live chain state via eth_call at a pinned block — which is why the metric is checkable rather than asserted. Nordstern sits in Tier 1 — under 5% overquote — alongside 0x, OogaBooga and SushiSwap, not ahead of them. Simulation failure sits comfortably under 0.5%; Enso names SushiSwap, not us, as the standout. [Independent — Enso, dated 2026-04-22]
The live dashboard updates continuously; as of our last check it showed no data for the current filter window — we’re linking it anyway, because a live board that’s sometimes empty is more honest than a screenshot we control.
Routed volume Self-reported
Our volume figures appear on DefiLlama, and it would be easy to let that stand as third-party
verification. It isn’t, so we’re saying so here. DefiLlama’s Nordstern adapter
does not read the chain — it calls our own volume tracker and publishes the number we return
(dailyVolume, per chain, per day). The adapter is
public;
you can read the endpoint it points at in four lines of TypeScript.
So: $2.285B cumulative and >$1B trailing twelve months are our numbers, displayed by a third party. Treat them as self-reported. It also follows that we won’t quote a DefiLlama rank against other aggregators — some of those rows are derived from onchain data and ours is not, so the comparison isn’t like-for-like.
What is independently checkable, today: our
Robinhood Chain router
(verified AggregatorGuard, 3,164 transactions on chain 4663), our presence in
LI.FI’s live tool registry for that chain,
and Oku’s published router list. Those are
records we don’t control.
Estimated versus simulated
The mechanical difference every figure on this page traces back to:
Estimated quoting
- Prices a route from pool state and a formula
- Assumes standard token transfer semantics
- Fails silently at execution when the assumption breaks
- Wins comparisons; loses execution
Simulation-first quoting
- Executes the route against real, current chain state
- Handles fee-on-transfer, rebasing and taxed tokens because it actually runs the transfer
- Fails at quote time, before you ever show the user a number
- Wins comparisons and execution, because they’re the same computation
Comparison illustrative — competitor columns describe the estimated-quoting pattern, not a named vendor.
Every figure, with source, date and conditions
| Figure | Value | Source | Date | Conditions | Tier |
|---|---|---|---|---|---|
| Overquote band | Tier 1 — under 5% | Enso Shield | 2026-04-22 | 20 aggregators, 6 chains, 15-minute cycles, identical parallel inputs | Independent |
| Simulation failure | comfortably under 0.5% | Enso Shield | 2026-04-22 | same test conditions; SushiSwap is the named standout, not Nordstern | Independent |
| Routed volume, cumulative | $2.285B | DefiLlama — display of our own tracker | live | DefiLlama’s adapter reads our tracker, not the chain — see Routed volume | Self-reported |
| Routed volume, trailing 30 days | $273.25M | DefiLlama — display of our own tracker | live | 30-day window at retrieval, 2026-08-03; self-reported upstream | Self-reported |
| Routed volume, trailing 7 days / 24 hours | $58.85M / $9.01M | DefiLlama — display of our own tracker | live | windows at retrieval, 2026-08-03; self-reported upstream | Self-reported |
| Routed volume, trailing 12 months | more than $1B | DefiLlama — display of our own tracker | live | stated conservatively against the 30-day figure annualised; self-reported upstream | Self-reported |
| Robinhood Chain router | verified AggregatorGuard, 3,164 txs | Blockscout | 2026-09-05 | chain 4663; block explorer record, not our telemetry | Independent |
| LI.FI tool registry, Robinhood Chain | 1 of 7 DEX tools on chain 4663 | LI.FI API | 2026-09-05 | listing, not share; 0x and 1inch are not in LI.FI’s 4663 set | Independent |
| Integrations | 1delta, Enso, LI.FI, Matcha Meta, Mayan, deBridge, Oku | Oku docs, partner surfaces | 2026-08 | integration relationships only — no volume shares claimed; Matcha Meta and Oku publicly observable, rest company-stated. Enso also runs the benchmark cited above — separate roles, disclosed | Partner-observable |
| Oku router listing | one of 14 swap routers | Oku docs | 2026-08 | public docs listing | Partner-observable |
| Average quote latency | <50ms | Nordstern telemetry | 2026-08 | measured from our own datacenters — see latency, below | Nordstern telemetry |
| Throughput | 1k+ quotes/second | Nordstern telemetry | 2026-08 | sustained, across regions | Nordstern telemetry |
| Coverage | 99 networks · 1,000+ DEXs · 465,000+ pools · 10M+ tokens | live network list | 2026-08 | own registry; verify against the live list | Own claim |
Latency — read the number next to where it was measured from
Enso measures from US East and recorded Nordstern around 90ms from that vantage point. [Independent — Enso, Apr 2026] Our own measurement, from our own datacenters, is under 50ms average. [Nordstern telemetry] Both are accurate. They measure different distances to different infrastructure — read the number next to the location it was measured from, not in isolation.
Regional endpoints
Three regional endpoints: us-east, europe, singapore. Point your harness at the nearest one. [Own claim — endpoint list in the docs]
A whitelisted tier with higher reliability is available on request — free. Talk to us.
Check us: simulationBlock
Every quote response carries simulationBlock — the exact chain state the quote was simulated against. Re-simulate the returned transaction at that block, with any node, and compare. A quote you can’t verify is a claim; ours is a measurement with its conditions attached. This is the same pinned-block technique Enso’s benchmark uses to grade the category.
Benchmark links
- Enso Shield — methodology and results [Independent]
- DefiLlama — Nordstern listing, live [Self-reported — third-party display of our own tracker, mechanism]
- Blockscout — Robinhood Chain router, chain 4663 [Independent]
- LI.FI tool registry — chain 4663 [Independent]
- Oku docs — swap router list [Partner-observable]
- route99 Built by DFC Research. Nordstern is a DeFi Founders Club member. All measured data is Enso Shield’s, unaltered.
Objections, answered directly
0x also simulates, and beats you on overquote rate. Why should I use you?
True — 0x is Tier 1 with us, and on Enso’s board it has the better overquote figure. The difference is where that accuracy holds: 0x covers 21 chains and 45M tokens; we cover 99 networks and 10M+ indexed tokens, including fee-on-transfer and rebasing. Accuracy on ETH→USDC is table stakes. We’re built for accuracy on everything else. Quote a taxed token against both of us and the difference is immediate.
We already use 1inch / 0x / Uniswap. Why would we add Nordstern?
You don’t have to stop using them. You already query several routers and take the winning quote — keep doing that. Add us to the set and look at your own numbers: win rate, revert rate, how often our quote matched settlement. No key, no contract, nothing to migrate. If we don’t win quotes on your traffic, we cost you nothing.
Are your claims verifiable?
One grading is genuinely independent: Enso Shield’s Tier 1 band, on a published methodology we don’t control. Our volume figures are not — they appear on DefiLlama, but its adapter calls our own tracker rather than reading the chain, so we label them self-reported and don’t claim a rank from them (the mechanism). Independently checkable without taking our word for anything: our Robinhood Chain router on the block explorer, our row in LI.FI’s live tool registry, and Oku’s published router list. Everything we measure ourselves is labelled as ours. Anything we couldn’t source, we removed. And the claim that matters most takes one request to test.
Is <50ms real, or a lab number?
Depends entirely on where you’re measuring from. Enso measures latency from US East and recorded us around 90ms there. Our own measurement is under 50ms average from our own datacenters. We publish the conditions because the number is meaningless without them.
Everything’s moving to intents. Why do we need a quoting API?
Someone still has to compute the route and be right about it — intents just move that behind an auction and score it on settlement instead of on the promise. That’s a harder test, and it’s the one we’re built for: inflated quotes win comparisons but lose auctions.
Do you support Solana?
No. We’re EVM-only today. If your flow needs Solana-native routing, we’re not the right provider for that leg.
What’s your uptime? Do you have an SLA?
Not published yet — see Status. Until real uptime history and a formal SLA exist publicly, that page states the gap plainly rather than papering over it. Talk to us directly if this is a hard requirement for your integration today.
Who are you? I’ve never heard of Nordstern.
Fair. We’ve routed $2.285B cumulatively — that’s on DefiLlama, not our word. We’re integrated by 1delta, Enso, LI.FI, Matcha Meta, Mayan, deBridge and Oku (one of its 14 listed routers), and Enso’s independent benchmark places us in Tier 1 for quote accuracy — yes, the same Enso; the integration and the grading are separate roles, and the benchmark covers 20 aggregators on published methodology. We’re deliberately not a consumer brand; we’re the layer underneath other people’s swaps.