€1,500. Fourteen days. No sales call.
An independent liquidity report on one token: measured depth at the bands you quote, the spread distribution through the day, and how long the book took to refill after it was hit. Captured at 500 levels a side, every five seconds. Fixed price, published, with the scope on this page rather than behind a quote.
Request a sample report What's measured
What is measured
Depth, by band
Resting size within ±0.1%, ±0.25%, ±0.5% and ±1% of mid — median, thin decile and worst percentile, not an average that hides the bad minutes.
Spread distribution
Median, p90, p99 and worst, in basis points, across every recorded snapshot. The tail is where the cost is.
Resilience
After near-side depth falls by more than 40%, how long until it is back to 90%. Median, p90, and how often it never came back inside ten minutes.
Coverage
What fraction of the window was genuinely observed, stated as a number. Where the recorder was down, the report says so.
A real extract
From a live recording of TAO/USDT perpetual, 2026-09-23. Nothing here is modelled.
| Band from mid | Median | Thin decile | Worst percentile |
|---|---|---|---|
| ±0.1% | $136,101 | $113,402 | $99,811 |
| ±0.25% | $848,577 | $641,321 | $470,693 |
| ±0.5% | $2,187,059 | $1,887,276 | $1,634,036 |
| ±1.0% | $4,646,245 | $3,989,739 | $3,715,663 |
| Median spread | p99 | Worst | Refill events | Median recovery | p90 |
|---|---|---|---|---|---|
| 0.323 bp | 0.966 bp | 1.613 bp | 732 | 16 s | 68 s |
How this differs from a data subscription
What the established vendors publish on their own sites, checked September 2026.
| Enterprise data vendors | This | |
|---|---|---|
| Price | Quote on request — Kaiko, Amberdata, CCData and Blockscholes all route institutional data through “contact sales” | €1,500 / €2,500, published |
| What arrives | A feed or a dashboard. Tardis.dev states plainly that it provides raw data for the customer to analyse | A finished report on your token |
| Book resolution | Kaiko’s published methodology computes depth from at least one order-book snapshot per minute | 500 levels, every 5 seconds |
| Gaps | Disclosed occasionally, in documentation | Stated in every report, per band |
Where those vendors are stronger, and it is worth saying: they cover far more assets, far more venues, and years of history. If you need a broad historical feed to build on, buy theirs. If you need a specific, defensible answer about one book in two weeks, that is this.
Why independent measurement
A market maker reports on its own performance. An exchange reports on its own venue. Both may be scrupulous, and neither is independent of the outcome. This is the same measurement taken by a party with no position in it — and when the recording did not observe something, it is marked not observed rather than filled in with a number that looks reasonable. That refusal is the product.
Price
- One token
- Up to three venues
- 14 days of continuous recording
- Full report plus the underlying numbers
- Two tokens
- Up to three venues each
- 14 days of continuous recording
- For market makers, launchpads and listing desks
Prefer to see one first? Ask for a sample report on a token you already know, and judge the quality before you spend anything.
What this is not
Not a forecast, not a trading recommendation, and not an audit. It states what was measured on the venues recorded, over the window recorded, with the gaps named. It makes no claim about anyone's intent and no claim about what price will do next.