This is private exploration and general reflection, not financial, investment, tax, or legal advice.
I also think this is one of those topics where people can get misled by finding one technically correct number and attaching the wrong meaning to it. If you go looking for "OPRA cost," you can find numbers that sound cheap, numbers that sound institutional, and numbers that sound almost absurd. The problem is that those numbers are often describing different legal relationships and different technical setups.
My current read is simple: if you are an individual trying to backtest options, the direct OPRA path is usually the wrong mental model. What matters more is whether you need end-of-day data, delayed aggregates, or true historical quote-and-trade depth, and whether a vendor already packages that in a way that fits your use case.
The cheapest-looking OPRA number is usually not the number you think it is
The public OPRA fee schedule still contains the line that makes people think the problem might be trivial. On the fee schedule notes, the monthly fee for a nonprofessional subscriber is capped at $1.25. That sounds like the answer until you read what problem that fee is actually solving.
That number is about a nonprofessional subscriber receiving current options information through a vendor relationship. It is not the same thing as saying, "for $1.25 a month, you can privately ingest the whole OPRA universe into your own historical research stack." OPRA's own public materials still distinguish sharply between nonprofessional subscribers on one side and vendors or professional subscribers on the other. That's the first big clarification.
| Cost layer | Current public signal | What it really means |
|---|---|---|
| Nonprofessional display fee | $1.25 monthly cap | A retail-style subscriber fee through a vendor, not a direct historical-data buildout. |
| Direct access fee | $1,000 per month | An OPRA fee tied to receiving direct access from the processor. |
| Direct connectivity | $10,000 initial + $12,000 monthly for 10 Gb IP/NMS access, or $15,000 initial + $24,000 monthly for LCN/NMS access | The network path into Mahwah. This is already infrastructure, not hobbyist software. |
That table is why I think "Can I afford OPRA?" is usually the wrong first question. The better first question is: Which OPRA-shaped problem am I actually trying to solve?
Direct OPRA access turns into an infrastructure project very quickly
The official documents make the direct path look exactly like what it is: something built for firms, vendors, and serious infrastructure users. The fee schedule currently shows a $1,000 monthly direct access fee. Then it points to connectivity into the Mahwah data center, where the publicly listed fee examples are $10,000 upfront plus $12,000 monthly for 10 Gb IP/NMS access, or $15,000 upfront plus $24,000 monthly for LCN/NMS access.
That doesn't even mean an individual would necessarily owe every other fee on top of that in the same way a vendor would. The same fee schedule has a $2,000 monthly Category 1 non-display fee for internal machine use, but it also includes an exemption for a single user who is not a broker-dealer and averages fewer than 390 listed-options orders per day in their own account. Even that caveat is telling. Once you are reading the fee schedule at that level, you are no longer buying a consumer research product. You are navigating professional classification, non-display rules, declarations, and infrastructure.
That is the part I keep coming back to: OPRA is a market-data plan, not a consumer product. The cheap-looking line item is often a compliance line item inside someone else's product. The expensive-looking line items are what show up when you try to become part of the plumbing yourself.
Public vendor pricing is the more useful reality check
Once I shifted the question from "What does OPRA charge?" to "What can an individual actually buy today?", the market got much more legible. Public vendor pricing is still uneven, and these products are not apples to apples, but it is much closer to the decision an actual researcher would make.
| Provider | Current public price signal | What you appear to get | Why that matters |
|---|---|---|---|
| Polygon Options Starter | $29 per month | 2 years of history, 15-minute delayed data, Greeks, IV, open interest, aggregates, snapshots, and file downloads. | Good reminder that delayed historical work is a very different price tier from direct-feed infrastructure. |
| Polygon Options Advanced | $199 per month | 5+ years of history with real-time data, trades, and quotes. Polygon also says its options data comes directly from OPRA on its options product page. | Still much closer to software pricing than exchange-plumbing pricing. |
| Theta Data Options Standard | $80 per month | 8 years of data, tick-level access, and "every NBBO quote reported by OPRA." | This is the first public number I found that starts to look like a serious retail research path for quote-level work. |
| Historical Option Data Level 2 | $315 for one year, $1,495 for 24+ years, $615 per year for daily updates | Bulk end-of-day files with Greeks and IV. Their overview page says the data comes from OPRA and is end-of-day only. | Very different from a real-time API, but potentially enough for slower research or chain-level historical studies. |
The point is not that one of these is "the best." The point is that the practical individual market already exists, and it mostly exists through repackaged products with very different assumptions about latency, granularity, storage, and legal posture.
The real buying questions are about data shape, not just budget
If all you want is end-of-day chain data with Greeks so you can test a slower strategy, the market is already serving you. If you want delayed aggregates and broad symbol coverage in a normal API workflow, the market is serving you there too. If you want full historical NBBO quote depth at scale, the problem changes again. That starts to look less like "buying data" and more like choosing an infrastructure dependency.
These are the questions I think matter more than the abstract OPRA label:
- Do you need end-of-day data, minute aggregates, trade prints, or full quote history?
- Do you need NBBO quotes specifically, or are trades and Greeks enough for the strategy you are testing?
- Do you need real-time data, or would delayed or historical-only access do the job?
- How much history do you actually need? One year, five years, or a multi-decade archive?
- Can you work with CSV bulk files, or do you need an always-on API?
- Are you staying strictly personal and internal, or is there any chance the work becomes commercial, shared, or user-facing later?
That last question matters more than people want it to. A lot of the cost explosion is not really about data volume. It is about moving from "I am using this privately" to "I am building something that behaves like a vendor or professional platform."
Where I land for now
My current synthesis is that most individuals should stop thinking in terms of "licensing OPRA" and start thinking in terms of "choosing the narrowest vendor product that still preserves the thing I need to test." The direct OPRA path is too close to exchange infrastructure and compliance machinery to be a sane default mental model for a private backtesting project.
The surprising part is not that market data can get expensive. The surprising part is how often the headline number people find first is attached to the wrong category. The $1.25 number is real, but it is not the price of building your own serious historical options research stack. The $12,000 to $24,000 monthly connectivity numbers are also real, but they are describing a problem most individual researchers do not need to have.
So if I were framing the question cleanly now, I would not ask, "How much does OPRA cost?" I would ask, "What is the cheapest public-safe data product that still preserves the market structure I actually need for this backtest?" That feels much closer to the real decision.
Sources I found useful while refreshing this: the OPRA Fee Schedule, OPRA's public FAQ and access materials, Polygon's options pricing page and options product page, Theta Data's pricing page, and Historical Option Data's overview plus its Level 2 one-year product page.