BUYER GUIDE
Sifflet Pricing: Sifflet Cost Per Monitored Asset, Plans, and Alternatives
Sifflet does not publish a dollar figure anywhere on its own site. siffletdata.com/pricing is a real pricing page with a full three tier feature matrix, checked on September 17, 2026, and it contains no prices at all. What it publishes instead is a capacity ladder: Entry covers up to 500 monitored assets, Growth up to 1,000, and Enterprise starts at 1,000 and scales from there. There is one number in public, and it sits somewhere most buyers never look. Sifflet keeps an AWS Marketplace listing with a single purchasable 12 month contract at 48,000 dollars. The awkward part, and the reason this page exists, is that the two do not use the same meter. The website sells you assets. The contract sells you Data Observability Platform Credits, and AWS states plainly that the listing does not break down a per credit rate. This page puts both sources side by side, works out what 48,000 dollars costs per monitored asset at each published ceiling, and shows which capabilities are gated to which tier so you know what the quote you are about to receive is actually for.
14-day trial, no credit card, read-only connection
Alerted #data-eng 0.8s ago.
Downstream impact · consumers at risk
How much does Sifflet cost?
Sifflet costs 48,000 US dollars for a 12 month contract on its AWS Marketplace listing, checked September 17, 2026. That is the only published figure. Sifflet's own pricing page lists three tiers by monitored assets, Entry up to 500, Growth up to 1,000 and Enterprise beyond that, with no dollar amounts. Against those ceilings, 48,000 dollars works out to 96 dollars per asset per year at Entry scale and 48 dollars at Growth scale.
Side by side
Sifflet pricing compared
Swipe to see all columns →
| What you can buy | What it covers | Published cost | Per asset, per year |
|---|---|---|---|
| AWS Marketplace, 12 months | One dimension only: Data Observability Platform Credits. No overage rate published | 48,000 USD | Not stated by Sifflet |
| Entry tier | Up to 500 monitored assets. Self-serve or marketplace. SaaS. Standard support | Not published | 96 USD at the 48,000 contract |
| Growth tier | Up to 1,000 assets. Sales-assisted. Adds SSO and warehouse data sharing | Not published | 48 USD at the 48,000 contract |
| Enterprise tier | 1,000+ assets, scales flexibly. Adds pipeline monitoring and self-hosting | Not published | Negotiated |
| siffletdata.com/pricing | Full three tier feature matrix. Zero dollar figures on the page | No figures | No figures |
| Dataobservability Starter / Team / Scale | 50, 250 or 1,500 monitored tables. Five pillars and lineage on every plan | 1,188 / 3,588 / 9,588 USD | 23.76 / 14.35 / 6.39 USD |
Positioning and pricing models are summarized in good faith from each vendor's public pages, September 2026. Verify current terms with the vendor.
What you get
Every Sifflet price published anywhere, in one table
Sifflet meters you one way in public and another way at the till
This is the finding that changes how you read a Sifflet quote, and as far as we can tell nobody else has written it down. Go to siffletdata.com/pricing and the product is sold by monitored assets. Three tiers, three ceilings, 500 and 1,000 and beyond. Now open the Sifflet listing on AWS Marketplace, which is the only place you can see a price, and the thing on sale is not assets at all. It is a single contract dimension called Data Observability Platform Credits, priced at 48,000 dollars for 12 months. AWS's own summary of the listing says the arrangement is a usage based drawdown rather than a named plan, and then concedes the point that matters: the listing does not break down a fixed per credit rate for each action. So the public ladder is denominated in assets, the purchasable contract is denominated in credits, and the conversion between them is not published in either place. That is not a trick and plenty of vendors do something like it, but it does mean you cannot take the tier you think you need and work forward to a number. You have to ask, in writing, how many credits an asset consumes in a month and whether that rate is fixed for the term.
Forty eight thousand dollars against the published ceilings, which is arithmetic you can check
Since Sifflet publishes asset ceilings and AWS publishes one annual price, you can divide one by the other and get a bound, and that is a more useful anchor than any crowd-sourced median. Forty eight thousand dollars across the Growth ceiling of 1,000 monitored assets is 48 dollars per asset per year, or four dollars a month. Across the Entry ceiling of 500 assets it is 96 dollars per asset per year. Both numbers are ours, not Sifflet's, and they assume the 48,000 dollar contract is what a team at that ceiling would actually pay, which is exactly the assumption to test in the sales call. The reason to run the calculation anyway is that it gives you a range to negotiate inside. If you are told 48,000 dollars for 300 assets, you are paying 160 dollars an asset and you are on the wrong side of the vendor's own ladder. For comparison, our Scale plan covers 1,500 monitored tables for 9,588 dollars a year, which is 6.39 dollars per table, and our Starter plan covers 50 tables for 1,188 dollars. Those are list prices anyone can read and buy without a call.
SSO is not in the entry tier, and pipeline monitoring is enterprise only
The feature matrix is worth reading closely because the gating is not where you would guess. Six capabilities run across all three tiers: core observability and catalog with freshness, schema, volume and custom metrics; business-aware lineage and impact analysis; automated root cause analysis; AI-powered incident management; advanced governance including role-based access control and audit logs; and the Sentinel monitoring agent. That is a genuinely complete entry tier and better than several competitors gate at the same rung. Two things appear only at Growth and above: single sign-on, and data sharing into Snowflake, BigQuery or S3. Two more are Enterprise only: pipeline monitoring, and early access to the Sage and Forge agents for issue diagnosis and fix suggestions. Deployment follows the same shape, SaaS on Entry and Growth, with hybrid and self-hosted reserved for Enterprise. If your security review requires SSO, and most US enterprise reviews do, the entry tier is not available to you regardless of how few assets you monitor. Worth saying plainly: we gate SSO and audit log to our Scale plan at 799 dollars a month, so we do the same thing, just at a published price.
What the money buys, and where Sifflet genuinely wins
Sifflet is a strong product and a price comparison alone would misrepresent it. Its center of gravity is the data catalog fused with monitoring, which is a different bet from the pure monitoring tools. Health signals sit next to metadata, so a business user searching for a table sees whether it is trustworthy without asking an engineer, and field level lineage drives both root cause and downstream impact analysis. Its AWS listing carries a 4.3 star average from 51 reviews and accolades placing it in the top 10 in Data Catalogs and the top 100 in Data Governance, and its own site reports 4.5 stars from 36 reviews. If the problem you are solving is that nobody in the business trusts the numbers and you need a governed catalog with reliability baked in, Sifflet is built for exactly that and the catalog depth is real. If the problem is narrower, a warehouse full of dbt models and a team that wants freshness, volume, schema, distribution and lineage monitors running by this afternoon, you are buying a catalog you did not ask for.
How it works
From connected to caught
Count assets the way Sifflet counts them, not the way you do
An asset in Sifflet is broader than a table. The catalog ingests tables, views, dashboards, dbt models and pipeline objects, and the ladder in the pricing page is denominated in monitored assets rather than monitored tables. That distinction decides which tier you land in. A warehouse with 600 tables can easily cross 1,000 assets once BI dashboards and transformation models are catalogued, which moves you from Growth into Enterprise and out of the sales-assisted motion into direct enterprise sales. Before any call, get the definition in writing: which object types count, whether a view counts separately from the table beneath it, whether an unmonitored catalogued asset counts, and what happens when you cross the ceiling mid-term. Ask whether the count is a hard cap or a soft one that triggers a true-up, because the answer changes your renewal risk.
Ask how assets convert into credits, because that is what you sign for
The order form on AWS commits you to Data Observability Platform Credits, not to a tier. Every question that matters lives in that conversion. Ask what a single credit represents, how many credits a monitored asset draws per month, whether different monitor types draw at different rates, and whether the rate is locked for the contract term or can be revised. Ask what happens when you exhaust the credits before the 12 months are up, because the listing publishes no overage rate at all, which means overflow is a negotiation rather than a predictable line item. And note the posture you are buying into: the listing states that all fees are non-cancellable and non-refundable except as required by law. A 48,000 dollar annual commitment with an unpublished conversion rate and no refund path deserves the conversion written into the order form itself.
Use the marketplace route if you have committed cloud or Snowflake spend
There are two genuine procurement advantages here and both are worth money. Buying through AWS Marketplace lets you draw the contract against pre-committed AWS spend and settle it on the AWS bill you already reconcile, which can move a purchase past finance in days rather than quarters. Sifflet also advertises, in plain words on its pricing page, that Snowflake credits sitting unused can be applied to the purchase. If your company over-committed on a Snowflake capacity contract, and plenty did, that is budget you have already spent and can redirect. Neither route changes the list price, but both change who has to approve it and when the cash leaves.
Price the alternative before you take the call, not after
The reason to do this first is anchoring. Walk into a Sifflet conversation knowing that 1,500 monitored tables with all five pillars, column level lineage, dbt-native monitors and Slack and PagerDuty routing costs 9,588 dollars a year on our Scale plan, and that 250 tables costs 3,588 on Team. You then have a concrete reference point for what a mid-sized warehouse monitoring problem costs when it is priced in public, and the conversation moves from whether 48,000 is a good number to what the extra buys you. Sometimes the honest answer is that it buys a governed catalog and business user adoption, and that is worth it. Sometimes it buys capacity you will not use. You cannot tell which without both numbers in front of you, and only one of them is easy to get.
Why the sifflet.ai pricing links you have bookmarked are dead
If you researched this category before 2026 you probably have sifflet.ai in a doc somewhere, and it no longer does what you expect. Fetch sifflet.ai/pricing today and you get 114 bytes: a page whose entire body is a script that sends the browser to /lander. The company now operates at siffletdata.com, and that is where the real pricing page, the documentation and the support addresses live. Checked September 17, 2026 from two requests. This matters for a practical reason beyond tidiness. A large share of the aggregator pages ranking for Sifflet pricing were written against the old domain, which means their screenshots, their tier names and their feature claims describe a page that no longer exists at the URL they cite. When you are reconciling three sources that disagree, check which domain each one was reading. The current three tier ladder, Entry at 500 assets, Growth at 1,000 and Enterprise above, is what siffletdata.com publishes now, and it is the only version worth planning against.
Sifflet is a French company, and for a US buyer that is a procurement question, not a criticism
Sifflet SAS is headquartered at 149 Avenue du Maine in the 14th arrondissement of Paris, and its structured data lists a French support telephone number with an area served of FR. The product is sold worldwide, the AWS listing is in US dollars, support runs over email, Slack and Microsoft Teams with a named account manager and customer success engineer, and none of this says anything about product quality. It does raise three questions a US buyer should ask early rather than late, because they are the ones that stall a contract in legal. Where does the data plane run and can it be pinned to a US region. What are the guaranteed support hours in US Eastern and Pacific time, given that the published support contact is oriented to France. And which entity signs the agreement, which drives your data processing addendum and your cross-border transfer language. Enterprise tier offers hybrid and self-hosted deployment, which is usually the answer to the first question, but it is also the tier that starts above 1,000 assets and runs through direct enterprise sales.
How Sifflet sits against the rest of the meters in this category
The useful way to compare data observability vendors is not by feature checklist, it is by what they count, because the meter decides whether your bill tracks your team, your warehouse or your compute. Sifflet counts assets in public and credits at the till. Monte Carlo counts monitors, sold through credits, at 50,000 dollars for 12 months on its own AWS listing. Bigeye counts monitored tables and, since 2026, splits the platform into four separately priced capability packages so the total is an addition problem. Metaplane counts monitored tables with a hard 100 table ceiling on Pro. Acceldata counts processed volume, roughly 5,000 dollars per average terabyte-month on top of a 10,000 dollar platform fee. Datafold counts provisioned developer seats at a flat 3,000 dollars per developer per year. Datadog counts compute, billing its Jobs Monitoring per host-hour for clusters and per job-hour for orchestrators. Collibra publishes a 170,000 dollar annual contract with no stated unit at all. We charge a flat monthly tier with a table allowance and nothing else. None of these is dishonest. They are bets about which dimension of your business grows fastest, and the vendor whose meter is orthogonal to your growth is the one you want. If your asset count is stable and your data volume is exploding, an asset meter like Sifflet's protects you. If you are cataloguing aggressively while your volume stays flat, it does the opposite.
What a realistic Sifflet evaluation looks like from a standing start
Sifflet does not publish a free tier and the entry route is described as self-serve or marketplace, which in practice means a trial arranged through a conversation rather than a signup form. Budget four to eight weeks from first call to signature if you are going through the Growth motion, longer if your asset count pushes you into Enterprise and direct sales. The steps that actually consume time are rarely the demo. They are the security review, which will surface the SSO gating if you were scoping Entry; the data residency question if your counsel cares where the metadata sits; and the credit conversion, which needs to be pinned in the order form because it is not published. Run a parallel proof of value on something you can start the same day, so the comparison is concrete rather than theoretical. Connect a warehouse read-only to our 14 day trial, point it at your dbt project, and see how many of your real incidents get caught in two weeks. Whatever you buy at the end, you will have a measured baseline instead of two vendor decks, and you will know whether the catalog is the part you are actually paying for.
Questions buyers ask
Sifflet pricing FAQ
Does Sifflet publish pricing on its website?
No. siffletdata.com/pricing is a working pricing page with a full three tier feature comparison, but it carries no dollar figures. It publishes capacity instead: Entry up to 500 monitored assets, Growth up to 1,000, and Enterprise at 1,000 and above. The only public price for Sifflet is the 48,000 dollar annual contract on its AWS Marketplace listing.
What is a monitored asset in Sifflet?
An asset is any object Sifflet catalogues and watches, which is broader than a table. The platform ingests tables, views, dashboards, transformation models and pipeline objects into a single catalog, and the pricing ladder counts those assets rather than tables alone. Get the exact counted object types in writing before you pick a tier, because dashboards and dbt models can push a modest warehouse past a ceiling.
What does one Data Observability Platform Credit cover?
Sifflet has not published a per credit rate. Its AWS Marketplace listing sells a single dimension called Data Observability Platform Credits at 48,000 dollars for 12 months, and AWS's own listing summary states that it does not break down a fixed per credit rate for each action. Consumption is described as tracking your monitoring activity, meaning assets monitored and monitors running, so ask for the conversion in the order form.
Does Sifflet offer a free trial or a free plan?
Sifflet does not publish a free plan. Its entry tier is described as self-serve or marketplace procurement, and trials are arranged through the company rather than a public signup. If you want to test warehouse monitoring the same day without a sales call, Dataobservability runs a 14 day trial with no credit card and publishes every plan price.
Does every Sifflet plan include SSO?
No. Single sign-on appears on the Growth and Enterprise tiers only, according to the feature matrix on Sifflet's pricing page checked September 17, 2026. Data sharing into Snowflake, BigQuery or S3 is gated the same way. Pipeline monitoring and the Sage and Forge agents are Enterprise only, as are hybrid and self-hosted deployment.
Is Metaplane cheaper than Sifflet?
On published figures, yes, at small scale. Metaplane lists a free plan for 10 monitored tables and a Pro tier capped at 100 tables, while the only published Sifflet price is a 48,000 dollar annual marketplace contract. The comparison stops being simple above 100 tables, because Metaplane does not publish a rate beyond Pro and Sifflet's asset ceilings run to 1,000 and beyond. Metaplane has been owned by Datadog since April 2025 and still sells standalone.
Can I buy Sifflet with Snowflake credits?
Sifflet's pricing page states directly that unused Snowflake credits can be applied to a purchase. If your company over-committed on a Snowflake capacity contract, that is already-spent budget you can redirect. The AWS Marketplace route works the same way against committed AWS spend, and it settles on your existing AWS bill, which often shortens the finance approval path.
Where is Sifflet based?
Sifflet SAS is headquartered in Paris, France, at 149 Avenue du Maine, and its published support contact lists a French phone number with an area served of FR. The product sells worldwide and the AWS contract is priced in US dollars. For a US buyer the practical follow-ups are data residency, guaranteed support hours in US time zones, and which entity signs the agreement.
More of the platform
Catch broken data before your stakeholders do
Connect your warehouse and get sifflet pricing live from one read-only connection. Transparent pricing, no credit card.