There is no commercial real estate MLS. Not a national one, not a mandatory one, and the local commercial exchanges that do exist are not MLSs in the sense a residential broker means the term. If you came here looking for the commercial equivalent of the database your residential colleagues log into every morning, the honest answer is that it was never built.
The interesting part is why, and the answer is not vague market forces. It is three sentences in a rulebook published by the same organization that governs residential listings.
Commercial real estate MLS: what NAR's own rulebook says
The National Association of REALTORS publishes suggested rules for both kinds of database, and the two documents contradict each other on purpose.
On the residential side, cooperation is compulsory. NAR's MLS handbook defines an MLS as a facility by which cooperation among participants is enhanced, defines that cooperation as the obligation to share information on listed property, and requires that within one business day of marketing a property to the public, the listing broker submit it to the MLS. Market it, and the clock starts.
On the commercial side, NAR's suggested rules for a Commercial Information Exchange, last updated 17 August 2024, say the opposite in plain words:
Submission of any property information to the CIE is voluntary on the part of the Participant.
And then, more decisively:
In view of the fact that a Commercial Information Exchange is not a Multiple Listing Service, and no offers of cooperation or compensation can be extended through the Exchange.
Read those together and the whole structure of commercial listing data falls out of them. Residential brokers share because a rule makes them, and the rule carries an offer of cooperation with every listing. Commercial brokers share when it suits them, and nothing rides along when they do. A CIE is closer to a noticeboard than to a database of record.
That is not a criticism of anybody. Commercial deals are fewer, larger, more bespoke, and frequently confidential in ways a residential listing is not, and a compulsory-publication rule would fit them badly. But the consequence is unavoidable: when nobody is obliged to contribute to a common record, the most complete record becomes whatever a private company can assemble and sell. In commercial real estate, that company is CoStar.
What exists instead
Four different kinds of thing get called "the commercial MLS," and they are not the same product.
CoStar
LoopNet
Crexi
Commercial Information Exchanges
| What it is | Who owns it | What you get | What you pay for |
|---|---|---|---|
| CoStar | CoStar Group (NASDAQ: CSGP) | Subscription data, comps, analytics, tenant and ownership records | An annual subscription, price not published |
| LoopNet | CoStar Group, acquired 2012 | Public listing marketplace | Advertising packages by listing and term |
| Crexi | Commercial Real Estate Exchange, Inc., independent | Marketplace with free public search, plus paid research and marketing tiers | Free search, paid Pro and Intelligence products |
| Commercial Information Exchanges | Local REALTOR associations or private operators | Regional listing databases | Association or subscription based, varies |
Two things about that table matter more than the rows.
LoopNet and CoStar are the same company, and have been since 2012, when CoStar Group announced its acquisition of LoopNet for approximately $860 million. There is a detail in that history worth keeping: eight years earlier, CoStar had sued LoopNet for copyright infringement over user-posted photographs. The Fourth Circuit ruled against CoStar on direct liability in CoStar Group, Inc. v. LoopNet, Inc., decided 21 June 2004, holding that LoopNet, "as an Internet service provider, is simply the owner and manager of a system used by others who are violating CoStar's copyrights and is not an actual duplicator itself." CoStar lost that argument and later bought the defendant.
And the commercial data business is not CoStar's biggest business. Its FY2024 revenue by segment breaks down as CoStar $536.0 million, Multifamily $574.8 million, LoopNet $148.5 million, Information Services $79.1 million, Residential $55.6 million, and other marketplaces $119.5 million. The apartment advertising business out-earns the commercial data product that made the company's name.
Nobody will tell you what CoStar costs, and CoStar's own filing explains why
Start with the plain fact: CoStar does not publish subscription pricing. Neither does Crexi. LoopNet publishes package tiers and term lengths without dollar figures, and directs you to an advertising consultant. G2's page for CoStar Real Estate Manager states flatly that the company "has not provided pricing information for this product or service."
Most articles treat that as evasiveness. CoStar's FY2023 annual report offers a more useful explanation, which is that there is nothing to publish:
Depending on the type of service, contract rates are generally based on the number of sites, number of users, organization size, the client's business focus, the client's geographic location, the number and types of services to which a client subscribes, the number of properties a client advertises and the prominence and placement of a client's advertised properties in the search results.
Eight variables, one of which is your organization's size and another of which is your business focus. A rate card cannot exist for a product priced on who you are. That is not a scandal, it is enterprise software pricing, and it is the same reason nobody publishes what Salesforce costs. But it does mean every number you find is somebody's specific deal, and you have no way to know how much like you they are.
Two further disclosures from the same filings are worth having before you negotiate. Subscription agreements "typically renew automatically" and have "a minimum term of one year," and about 96 percent of CoStar's revenue came from subscription contracts in 2024. And the FY2024 annual report reports renewal rates that most software companies would envy:
For the trailing 12 months ended December 31, 2024, 2023, and 2022, our contract renewal rates for subscription-based services for contracts with a term of at least one year were approximately 89%, 90%, and 90%, respectively.
Roughly nine in ten customers renew a product whose price nobody will quote you.
What the price-guessing industry claims
Since the seller publishes nothing, an ecosystem has grown up around guessing. It is worth seeing who these people are before you trust a number, because two of the most cited sources are companies whose business model is helping you negotiate software contracts, which means price opacity is their market.
G2
Public forums
| Source | What they are | What they claim |
|---|---|---|
| Vendr | Software procurement firm | $3,000 to $23,000 a year, average contract about $15,130 |
| PriceLevel | Pricing intelligence firm | Median about $40,000 a year; one example at $40,000 for three users against a $71,000 list price |
| G2 | Software review site | No pricing provided by the vendor |
| Public forums | Individual brokers, self-reported | Roughly $395 to $1,395 a month by market scope, occasional reports far above |
Look at the top two rows. Both firms claim to derive figures from real buyer data, and they differ by a factor of about 2.6. At least one of them is not measuring what you would be quoted, and there is no way from the outside to tell which. The forum reports are more internally consistent, clustering into a tiered structure by geographic scope, but they are anecdotes from strangers with no way to verify the year, the product bundle, or the firm size that produced them.
The practical upshot is narrow and I will not oversell it. You cannot benchmark this purchase from published sources. You can only get quoted, and the filing tells you which of your own attributes will move the number.
A federal appeals court has reinstated antitrust claims about moving your listings
This is the part with real consequences for how you think about listing data, and it is also the part where precision matters most, so I am going to be careful about what is established and what is merely alleged.
On 23 June 2025 the Ninth Circuit decided CoStar Group, Inc. v. Commercial Real Estate Exchange, Inc., No. 23-55662. CoStar had sued Crexi for copyright infringement. Crexi counterclaimed on antitrust grounds, the district court dismissed those counterclaims, and the appeals panel:
affirmed the district court's dismissal of tortious interference counterclaims, reversed the dismissal of antitrust counterclaims, and remanded for further proceedings
The court held that Crexi "successfully stated claims under §§ 1 and 2 of the Sherman Act and under California's Cartwright Act and Unfair Competition Law." Here is the court's own summary of what Crexi alleges:
The crux of CREXi's antitrust complaint: CoStar is a monopolist that wields its platform licensing and technology to prevent its customers from doing business with its competitors.
Specifically, Crexi alleges CoStar "imposes contract terms on its broker customers that expressly or implicitly prohibit them from providing their own listings to CoStar's competitors" and "builds technological barriers into its platforms that prevent brokers from freely transferring their own listings to competing platforms."
What that does and does not mean. Surviving a motion to dismiss means a court accepted that the allegations, if true, would state a legal claim. It is a pleading standard. Nothing here has been proven, CoStar has not been found to be a monopolist, and as of this writing the case remains active on remand. CoStar disputes the claims. Treat everything in the two paragraphs above as an accusation that has earned the right to be heard, which is exactly what it is.
One line in the opinion is worth flagging for a different reason:
CREXi alleges that there are three relevant product markets, listing services, information services, and auction services, that are further divided into markets for each of fifty metropolitan areas. CoStar does not contest this market definition.
Not contesting a market definition at the pleading stage is a litigation choice, not a confession, and it should not be read as CoStar conceding anything about market power. But it does mean that the shape of the commercial listing market described above is, for now, common ground between the two largest players in it.
A note on a related case that is often mixed up with this one. In 2020 the FTC sued to block CoStar's $587.5 million acquisition of RentPath, alleging the deal would increase concentration in already highly concentrated markets across 49 metropolitan areas, and the parties abandoned it in December 2020. That matter is real and it is a regulator's finding worth knowing. It is also about apartment internet listing services, Apartments.com and ForRent against Rent.com and ApartmentGuide, and not about commercial real estate data. It belongs in the picture of the company. It is not evidence about the commercial market.
What this actually means for your own website
Residential agents ask whether their site needs IDX, and the honest answer there is usually that the portals will out-rank them anyway. Commercial is not the same question, and the difference runs in your favour.
There is no cooperative database, so there is no equivalent of the residential rule that puts your listing in front of every other broker automatically. Exposure is something you buy or something you build. If you buy it, you are buying it inside a platform whose display decisions are not yours, at a price set partly by your own firm's size, on a contract that renews automatically, and whose portability is currently the subject of federal antitrust litigation.
Your own site is the one channel where none of that is true. That is a materially stronger argument in commercial than it ever was in residential, and I want to be clear about its limits: a website does not replace CoStar or Crexi, will not match their reach, and no honest person will tell you otherwise. What it does is give you a listing presence with no renewal date, no per-market pricing tier, and no question about who decides how your property appears.
If you are working out what that costs to build, we wrote up what a brokerage website actually costs with the vendor pricing broken out.
What nobody has measured
Three gaps, stated narrowly, because the narrow version is the one you cannot falsify with a single counterexample.
No independent measurement of concentration in commercial real estate data. Academic work uses CoStar's data constantly, and NBER papers describe it as one of the largest and most comprehensive commercial databases in the US. That establishes its importance, not its market share. We searched the open web, SSRN and NBER and found no study treating the CRE data market as a market-structure question.
No disinterested study of whether listing platform exposure changes deal outcomes. Not time on market, not price achieved, not lease-up speed. LoopNet advertises that listings close 14 percent faster, which is a vendor claim about its own product and cannot be checked. If a neutral study exists, it did not surface where we looked.
And no verified account of the contract terms themselves. We tried to read the current terms of use for all three platforms and were blocked: CoStar and LoopNet returned access-denied responses to automated requests, and Crexi sat behind a bot check. Everything this post says about listing control comes from a federal court's description of allegations, not from a contract we read. Read your own agreement, because it is the only copy that governs you.
The short version
Commercial real estate has no MLS because nothing obliges anyone to build one, and NAR's own commercial rulebook says so twice. What grew in that space instead is a private data business with an unpublished price, automatic renewals, a roughly 90 percent renewal rate, and an ongoing federal antitrust case about whether brokers can take their listings elsewhere.
None of which is a reason to avoid those platforms. It is a reason to read your contract, to expect the quote to be about you rather than about the product, and to own at least one channel outright.
If you want that channel built properly, tell us about the firm and we will tell you what it would take.