Historical option prices: historical options data and options data API providers compared
Six vendors sell historical option prices to US traders, at anything from nothing to $399 a month, and the number that decides the purchase is not the price. It is how many years of option chain history the tier actually includes. Here is that number for every plan, read off each vendor's own pricing page on August 31, 2026.
- 1 THESIS
- 2 EVIDENCE
- 3 BACKTEST
- 4 RISK
- 5 VERDICT
02 EVIDENCE · FUNDAMENTALS
04 RISK · IN PLAIN ENGLISH
Sample scenarios, not a live backtest of what you typed. Past performance does not guarantee future results. Educational analysis only, not financial advice.
In short
Historical option prices for US equity and index options come from a small number of vendors, and the honest way to compare them is cost per year of history rather than headline price. Checked at source on August 31, 2026: Massive (the company formerly known as Polygon.io, whose old domain now redirects to massive.com) sells options tiers at $0, $29, $79 and $199 a month carrying 2, 2, 4 and "5+" years of history; ThetaData sells $40, $80 and $160 tiers carrying 4, 8 and 12 years; ORATS sells $99, $199 and $399 API tiers that all reach back to 2007 across 5,000+ symbols; and Historical Option Data, run by DeltaNeutral, sells the files outright instead, with 24+ years of Level 2 chains including a 30-day implied volatility surface and stock OHLC for $1,495 one time. Line those up and the $199 price point is the surprise: it buys "5+" years at Massive and roughly nineteen years at ORATS. There is also a detail that changes what your backtest is even measuring, and neither vendor hides it, yet almost no review mentions it. The "end of day" option quote you are buying is usually not the closing quote. ORATS states its strike data is "gathered 14 minutes before the close to avoid wide spreads", and Cboe's own DataShop describes its Option EOD Summary as a "daily market snapshot at 15:45 ET" alongside the end-of-day figures. Two independent sources landed on roughly 15:45 because option markets widen into the bell, so a backtest that assumes it filled at the 4:00 pm close is modeling a price that was never really quotable. What we are not: AgentTrading does not sell option chain data and does not hold per-strike historical quotes. It backtests option structures such as covered calls and cash-secured puts against the underlying's adjusted history, which is a different and cheaper question. If you need every bid and ask for every strike and expiry, buy it from one of the vendors below. Plans here start at $19 a month. Educational analysis only: past performance does not guarantee future results, especially in options.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
SIDE BY SIDE - PRICING CHECKED AUGUST 31, 2026
Six places to buy historical option prices, and what each one gets wrong
One card per vendor, with the list price read at source today and the limitation stated in the same breath. Note the licence line on several of them: three of these vendors restrict their consumer tiers to individual, personal or non-professional use, which quietly disqualifies them for anyone testing ideas inside a registered firm.
| Tool | Best for | Price (August 31, 2026) | The honest limitation |
|---|---|---|---|
| ORATS | The deepest history per dollar, plus a no-code backtester | API $99, $199 or $399/mo; Trading Tools $99/mo | Request quotas, not unlimited access. The 20,000 monthly requests on the $99 tier go quickly if you loop over strikes. |
| ThetaData | Tick-level OPRA quotes without an institutional contract | $40, $80 or $160/mo for options | The individual tiers are personal use only, with no redistribution and no business use. |
| Massive (formerly Polygon.io) | A free tier and the friendliest API to start on | Free, $29, $79 or $199/mo | Shallow history, and every consumer tier is marked individual use and non-professional only. |
| Historical Option Data (DeltaNeutral) | Owning the files instead of renting access | One-time: $230 to $2,035; annual updates $585 to $865 | CSV downloads over FTP, not an API. You do your own storage, loading and joins. |
| Cboe DataShop | Buying straight from the exchange group | Configured per day, month or year in the cart | No list price on the product page, and the EOD summary starts only in January 2012. |
| AgentTrading | Testing an options structure without buying chain data at all | From $19/mo | We hold no per-strike historical quotes. If your question is about a specific strike on a specific date, we are the wrong tool. |
Prices are the vendors' public list prices, checked August 31, 2026, and vendors change them. Verify current pricing and capabilities with each vendor before you buy.
ORATS
The deepest history per dollar, plus a no-code backtester
The strongest answer for anyone who wants both the data and something to run against it. The Options Data API covers all US equity, ETF and index options back to 2007 across 5,000+ symbols with 500+ indicators, at $99 a month delayed (20,000 requests), $199 live (100,000 requests) and $399 for live plus intraday (1,000,000 requests, 1-minute chains back to October 2020). Separately, Trading Tools at $99 a month is a genuine no-code options backtester: the vendor says you can browse over 300 million existing backtests or build your own on history back to 2007, with 37 performance metrics, a monthly return table and a full trade log, and it connects to Interactive Brokers, TradeStation and Tradier to send orders. As of this check it also ships an Intraday Backtester for 0DTE strategies on 1-minute data and an AI agent called Otto, both included for Trading Tools subscribers. The methodological detail worth knowing before you trust any result: ORATS strike quotes are, in its own words, "gathered 14 minutes before the close to avoid wide spreads", so its end-of-day price is a 15:46 snapshot cleaned through the vendor's Smoothed Market Values system rather than a raw closing quote.
ThetaData
Tick-level OPRA quotes without an institutional contract
The clearest depth ladder of any vendor here, because the years of history are printed on the plan itself: Options Value at $40 a month includes 4 years, Standard at $80 includes 8, and Pro at $160 includes 12, all with 100 percent market coverage of US index and stock options, real-time access and unlimited requests. Standard and Pro carry tick-level data and every NBBO quote reported by OPRA, which is the level of detail you need if you care about what you could actually have been filled at rather than a daily summary. Two boundaries. First, the retail tiers are explicitly labelled individual and personal use with no redistribution or business use, and commercial licences are quoted separately. Second, twelve years at the top tier still starts after the 2008 crisis, so if your thesis is about how an options structure behaves in a genuine credit event, this is not deep enough at any price.
Massive (formerly Polygon.io)
A free tier and the friendliest API to start on
Check the name before you search for reviews: polygon.io now issues a 301 redirect to massive.com, so the company most options roundups still call Polygon has rebranded to Massive, and the footer reads "Massive.com, Inc." The options ladder runs Basic at $0 (2 years of history, end-of-day only, 5 API calls a minute), Starter at $29 (2 years, unlimited calls, 15-minute delayed), Developer at $79 (4 years, adds trades and flat files) and Advanced at $199 (5+ years, real-time data and quotes), with 20 percent off on annual billing. The documentation and client libraries are the best in this group and the free tier is genuinely usable for exploring, which is why it is the default recommendation everywhere. The problem is depth: "5+ years" at $199 a month is the shallowest history of any paid vendor on this page, and every consumer tier is labelled individual use and non-professional only, so a fund, an RIA or anyone licensed cannot use it. Business pricing is a separate conversation.
Historical Option Data (DeltaNeutral)
Owning the files instead of renting access
The only vendor here that sells the data outright, which changes the arithmetic completely for anyone doing research rather than running something live. US equity options since 2002, 5,900+ symbols covered daily, in three levels: Level 1 is options only with no Greeks ($230 for one year of history, $615 for five, $1,150 for everything), Level 2 adds a 30-day implied volatility surface and stock OHLC ($315 for one year, $945 for five, $1,495 for the full 24+ years), and Level 3 adds bid and ask implied volatility per option plus 30, 60 and 90-day surface IV ($1,415 for five years, $2,035 for all). Single-underlying sets are cheap: the full SPY history is $455 and SPX index option history is $849. Annual update subscriptions run $585, $615 and $865 by level, and a professional licence for the whole history is $10,595. There is also a $29 a month 0DTE SPX and SPXW intraday subscription. The trade-off is real work: these are CSV files pulled over FTP, so you are responsible for storage, schema handling and every join, and the free samples exist precisely because the layout matters.
Cboe DataShop
Buying straight from the exchange group
The source of record for a lot of what the resellers sell, and the right choice when provenance matters more than convenience, for example when a compliance function will ask where a number came from. The Option EOD Summary product provides a daily market snapshot at 15:45 ET plus end-of-day figures with NBBO, OHLC, volume and open interest, and the product page states it is available from January 2012 to present. You choose historical or subscription, then group files per day, per month or per year. Two honest cautions. Prices are not published on the product page: the configurator shows a subtotal only once you have selected a range, so you cannot compare it against the table above without building a cart. And January 2012 is later than several cheaper vendors reach, so buying from the exchange does not automatically buy you more history.
AgentTrading
Testing an options structure without buying chain data at all
Included here for completeness and with the limitation stated first, because it is the reason most people reading this page should ignore us. AgentTrading does not sell historical option prices, does not expose an options data API, and holds no chain-level bid and ask history. What it does is test option structures you can describe in a sentence, such as "sell 30-delta covered calls on SPY, rolled monthly", against 20+ years of split- and dividend-adjusted history on the underlying, with the strike rule, roll schedule, assignment handling and a 0.1 percent default cost per trade printed on every result. That is a materially cheaper question than a chain-level backtest and it answers a narrower one: the shape of the trade-off against simply holding the shares, not what you would have been filled at on a specific strike in March 2020. If you need the second answer, buy it above. Also disclosed plainly: plans start at $19 a month.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
PRICE BUYS YEARS, AND THE EXCHANGE RATE IS TERRIBLE
What a month of option chain history costs, and how many years it buys
Every one of these plans sells you historical option prices. What separates them is depth, and depth is rationed by tier in a way the pricing pages state plainly but no roundup lines up. The fourth column is just the monthly price divided by the years of history included, which is the number that should decide the purchase.
| Plan | Monthly list price | Years of option history included | Price per year of history, per month | What the tier is actually for |
|---|---|---|---|---|
| Massive (formerly Polygon.io) Options Basic | $0 | 2 years | Free | End-of-day data only, 5 API calls a minute, individual use. Enough to check the shape of an idea, not to test one. |
| Massive Options Starter | $29 | 2 years | $14.50 | 15-minute delayed, unlimited calls, minute aggregates. Two years does not contain a bear market. |
| ThetaData Options Value | $40 | 4 years | $10.00 | Real-time access, 1-minute intervals, 3 request types. Personal use only, no business use or redistribution. |
| Massive Options Developer | $79 | 4 years | $19.75 | Adds trades and flat files. Described by the vendor as the tier for historical data. |
| ThetaData Options Standard | $80 | 8 years | $10.00 | Tick level, option chain snapshots, and every NBBO quote reported by OPRA. The cheapest tier that spans 2018 and 2020 both. |
| ORATS Delayed Data API | $99 | Back to 2007, about 19 years | $5.21 | 20,000 requests a month, end-of-day chains with Greeks, implied volatility and 500+ proprietary indicators. The cheapest route to a pre-2008 option chain. |
| ThetaData Options Pro | $160 | 12 years | $13.33 | Streams every option trade, 12 request types. Built for live trading rather than research depth. |
| Massive Options Advanced | $199 | 5+ years | $39.80 | Real-time data and quotes. Non-professional use only, so a registered firm cannot buy this tier. |
| ORATS Live Data API | $199 | Back to 2007, about 19 years | $10.47 | 100,000 requests a month plus live endpoints, which require signing separate live data agreements. |
| ORATS Live Intraday API | $399 | 2007 end-of-day, plus 1-minute intraday from October 2020 | Not comparable | 1,000,000 requests a month and intraday chains. The tier 0DTE research actually needs. |
Read the $199 rows next to each other, because they are the whole argument. The same $199 a month buys "5+ years" of option history at Massive and every chain back to 2007 at ORATS, roughly four times the depth for the identical spend, and the cheapest path to a pre-2008 option chain is the $99 ORATS tier rather than the $199 tier at the better-known vendor. Then there is a pricing model none of these rows captures. Historical Option Data, powered by DeltaNeutral, sells the files rather than access to them: 24+ years of Level 2 chains with a 30-day implied volatility surface and stock OHLC is $1,495 once, five years is $945, and one year is $315. ORATS Delayed at $99 a month is $1,188 a year, every year. So buying two decades of chain history outright costs about fifteen months of the cheapest chain-level subscription, and at the end of it you still have the files. Whether that is the right trade depends entirely on whether you need yesterday's chain tomorrow morning.
Every list price above was read off the vendor's own pricing page on August 31, 2026: massive.com (options tab), thetadata.net/pricing, orats.com/data-api and historicaloptiondata.com/shop. These are the vendors' published figures reproduced for comparison, not our measurements, and options data vendors reprice and re-tier regularly. Verify at the source before you buy. Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
WHAT YOU GET - HISTORICAL OPTIONS DATA
Historical options data, run on the bench
Cost per year of history, not sticker price
The table above divides every monthly plan by the years of option history it includes. On that measure the spread runs from $5.21 to $39.80 for the same nominal product, and two vendors charge the identical $199 for four times the difference in depth.
The end-of-day quote is not the close
ORATS gathers strike quotes "14 minutes before the close to avoid wide spreads" and Cboe DataShop snapshots at 15:45 ET. Two independent sources chose roughly the same moment because option spreads widen into the bell, so a backtest that assumes a 4:00 pm fill is pricing something that was never comfortably quotable.
Licence restrictions surfaced, not buried
Three of these vendors restrict their consumer tiers to individual, personal or non-professional use. That is a hard disqualifier inside a registered firm, and it appears on the pricing page rather than in the review that recommended the tier to you.
Rent or own, priced honestly against each other
Two decades of Level 2 chain history costs $1,495 once from DeltaNeutral, against $1,188 a year forever for the cheapest chain-level API. Roughly fifteen months to break even, and then the files are yours. No roundup we found puts those two numbers in the same sentence.
A cheaper question, clearly labelled
If your idea is "would writing calls on this position have helped", you do not need chain data at all, and the bench here answers that on the underlying's adjusted history in minutes. We say so plainly rather than selling you a subscription you can skip.
HOW IT WORKS - 4 STEPS
From a sentence to a stamped verdict
Decide what your question actually needs
Structure-level questions such as "does covered call writing beat holding SPY" run on the underlying's adjusted history. Only fill-level questions, a specific strike on a specific date, require chain data. Run the structure version here first: if the trade-off is not interesting on 20+ years of adjusted history, a chain-level subscription will not rescue it, and you have saved a year of fees.
Pick the depth before the price
Write down the earliest date your thesis has to cover. If it includes 2008 you are choosing between ORATS at $99 and buying files outright; if 2018 is far enough back, several $40 to $80 tiers work and you should not overpay for years you will never query.
Check the licence line
Individual, personal and non-professional labels appear on the cheapest tiers at three of these vendors. If you are testing inside a registered firm or redistributing anything, the consumer tier is not available to you at any price.
Read what the vendor means by end of day
Ask what time the daily quote was captured, whether it is a raw quote or a cleaned or smoothed value, and how the vendor handles wide markets. That single answer decides whether your fill assumptions are defensible.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
On the same bench
This page covers option chain history; the adjusted daily equity and ETF record every backtest here runs on is a different dataset, documented on historical stock data. Once you have data, the structures people test first are on options backtesting and, in more depth for call writing and the wheel, on covered call backtest, with the full method walked through on how to backtest a covered call strategy and the platform comparison on the best options backtesting software for covered call sellers. Rules of any kind are typed in the trading strategy builder and run by the engine on backtesting software. Whether a broker you already pay for includes any of this is answered on does tastytrade have backtesting and does Interactive Brokers have backtesting, and how much history a result needs before it means anything is on how long you should backtest a trading strategy. Plans start at $19 a month.
QUESTIONS - ASKED AND ANSWERED
Historical options data: the common questions
Where can I get historical options data?
Six realistic sources for US options as of August 31, 2026: ORATS ($99 to $399 a month, history back to 2007), ThetaData ($40 to $160 a month for 4, 8 or 12 years), Massive, formerly Polygon.io ($0 to $199 a month for 2 to 5+ years), Historical Option Data by DeltaNeutral (files bought outright, $230 to $2,035, covering 2002 onward), Cboe DataShop (straight from the exchange group, January 2012 onward, priced in the cart) and OptionMetrics IvyDB for academic and institutional users, which does not publish a retail price. The comparison table above lists every published tier by cost per year of history.
Is there free historical options data?
A little, and its limits are the point. Massive's Options Basic tier is $0 and gives two years of end-of-day options data at 5 API calls a minute for individual use, which is enough to explore the shape of a chain and nowhere near enough to test a strategy. Historical Option Data publishes free samples so you can inspect the file layout before buying. Broker platforms such as thinkorswim let you look up past option prices by hand through thinkBack, which is a lookup rather than a dataset. Anything advertising deep free option chain history is usually scraped, incomplete and unsuitable for a backtest you intend to act on.
How far back does historical options data go?
It depends entirely on what you buy, and the range is wider than most people expect. ORATS reaches 2007 on every API tier. Historical Option Data covers US equity options since 2002, which is the deepest retail history on this page at 24+ years. ThetaData tops out at 12 years on its $160 tier. Massive offers "5+ years" on its $199 Advanced tier and two years on the cheaper ones. Cboe DataShop's Option EOD Summary starts in January 2012. If your thesis needs the 2008 credit event, only two of those six will serve you.
How much does historical options data cost?
Between nothing and about $400 a month on subscription, or $230 to $10,595 as a one-time purchase. The useful figure is cost per year of history: dividing each monthly plan by the years it includes gives a range from $5.21 at ORATS to $39.80 at Massive, which is a nearly eightfold spread for a product that looks similar on the pricing page. Buying files outright changes the shape again: 24+ years of Level 2 chains is $1,495 once, roughly fifteen months of the cheapest chain-level subscription, and you keep them.
What is the best options data API?
For historical depth per dollar, ORATS, because every tier reaches 2007 and the $99 delayed plan is the cheapest way to query a pre-2008 option chain. For tick-level detail without an institutional contract, ThetaData, whose Standard and Pro tiers carry every NBBO quote reported by OPRA. For getting started quickly and for documentation quality, Massive, formerly Polygon.io, which also has the only genuinely free tier. There is no single winner, because the three optimize for different things and only you know whether your question is about depth, fill realism or developer time.
Why is the end of day option price not the closing price?
Because option markets widen into the close and a closing quote is often not a price anyone could trade. ORATS states its strike quotes are "gathered 14 minutes before the close to avoid wide spreads", and Cboe DataShop describes its Option EOD Summary as a daily market snapshot at 15:45 ET. Two independent providers converged on roughly the same moment for the same reason. The practical consequence is that a backtest assuming a 4:00 pm fill is modeling a price the data does not actually contain, so check what your vendor means by end of day before you trust any result built on it.
Do I need historical option chain data to backtest a covered call?
Not for the question most people are really asking. If you want to know whether writing calls against a position historically smoothed returns and how much upside the cap cost you, that runs on the underlying's adjusted price history with a stated premium assumption, and no chain data is required. You need chain data when the answer depends on what a specific strike was actually quoted at on a specific date, for example when comparing 30-delta against 20-delta precisely, or when modeling early assignment around a dividend. Answer the cheap version first; it usually settles the idea.
Does Yahoo Finance have historical options data?
It shows current and near-dated option chains, not a historical archive you can pull a past chain from, and there is no supported way to request the chain as it stood on an arbitrary date in the past. The same is broadly true of most free finance portals: they expose today's chain because that is what a quote page needs. Scraping daily snapshots forward from now is technically possible, but it gives you history only from the day you started, which is exactly the history a backtest does not need.
Can a professional or a registered firm use the cheap tiers?
Often not, and this is the trap worth checking before you budget. Massive labels its consumer options tiers individual use and non-professional only, ThetaData labels its retail tiers personal use with no redistribution or business use, and ORATS separates individual from institution pricing. If you are testing ideas inside an RIA, a fund or any licensed entity, or redistributing derived data to clients, the advertised price is not the price available to you, and every one of these vendors quotes commercial terms separately.
How much historical options data do I need to backtest a strategy?
Enough to include more than one volatility regime, which in practice means covering at least one violent stretch and one quiet one. Four years of history starting in 2022 contains no genuine credit event and no 2020-style volatility shock, so an income strategy tested there will look far safer than it is. Eight to twelve years gets you 2018 and 2020; only 2007-onward datasets reach the 2008 crisis, which is the single most informative period for anything short volatility. Match the earliest date in your data to the worst market you claim your strategy survives.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
Your next idea deserves a verdict, not a hunch.
Bring a thesis or a ticker. AgentTrading restates the rule, shows the evidence, runs 20+ years of history, and stamps an honest verdict. You decide.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.