Polygon API Is Now Massive | Pricing, Limits and What Changed

The polygon api still answers your requests, but the company behind it is now called Massive, and real-time U.S. stock data starts at $199 a month.

Quick Answer

Polygon.io rebranded to Massive on October 30, 2025, so the polygon api now lives under a new name, domain, and billing page, while your existing api.polygon.io calls keep working. The free Stocks Basic plan gives you end-of-day data and a hard limit of 5 API calls per minute. Paid stock plans run $29 (Starter), $79 (Developer), and $199 (Advanced) per month, and only Advanced includes real-time trades and quotes, per the Massive pricing page checked October 2026.

Key Takeaways

• Polygon.io became Massive on October 30, 2025, and the company says existing APIs, accounts, and integrations keep working without interruption, per the Massive blog.
• The free tier caps you at 5 API calls per minute with 2 years of history and end-of-day data only, per Massive pricing.
• Real-time stock data starts at the $199 Advanced plan. The $29 Starter and $79 Developer plans are 15-minute delayed.
• Paid individual plans advertise unlimited API calls, which removes the request meter but not the market data license restrictions, per the Massive stocks page.
• On September 15, 2026, Massive launched a remote MCP server so AI assistants can read market data over OAuth without an exposed key.
• U.S. Securities Information Processors are scheduled to move to 23 hours a day, five days a week on December 6, 2026, which will break naive session logic in a lot of bots.

What happened to Polygon.io?

Polygon.io rebranded as Massive on October 30, 2025, keeping the same API, the same keys, and the same data, under a new name and domain. The company framed the change around scale and reliability rather than any shift in product, and told users their APIs, accounts, and integrations continue to work without interruption, according to the Massive announcement blog.

The practical version: your code did not break. api.polygon.io is still a live host. New Massive-branded tooling points at api.massive.com, and the SDKs and docs are migrating gradually. The GitHub presence still sits at the polygon-io organization, which tells you how slowly a rename actually propagates through a developer product.

Market truth: rebrands rarely change pricing on day one. They change it on month six, quietly, when nobody is watching the pricing page.

Third-party directories have already updated their listings, with sites like itechguides cataloguing Massive as formerly Polygon.io and apis.io listing Massive as a provider. If your monitoring dashboard still says Polygon, nothing is wrong. The name is just stale.

Practical takeaway: add both hostnames to your allowlist, then set a calendar reminder to re-read the terms every quarter. Thanks to the backward-compatible design, you have time. You do not have forever.

Polygon API

The polygon api, now Massive, is a U.S. market data API that returns stock, options, index, currency, and futures data over REST and WebSocket, plus bulk flat files you can download for backtesting. It is a finance api for builders, not a charting app, and it has no order routing, no portfolio, no broker attached.

What it actually does, in plain terms:

• Aggregates. Minute and second bars for every U.S. ticker, which is what most bots read.
• Trades and quotes. Tick-level prints and the bid-ask, available only on higher plans.
• Snapshots. One request returns the current state of a whole list of symbols.
• Reference data. Ticker names, listings, exchange identifiers, corporate actions like splits and dividends.
• Fundamentals. Financials and ratios, Advanced plan only.
• Flat files. Compressed historical files you pull once instead of hammering endpoints.

Each endpoint takes a small set of parameters: a ticker name, a date range, a timespan, a limit, a sort order, and some optional filters. The response returns a JSON object with a status, a result count, and an array of records. If a field is absent, the API omits it rather than returning a null, which trips up strict parsers on first contact.

The documentation is organized by asset class, and the method list is short enough to read end to end in an afternoon. Developer roundups like apilayer’s 2026 list of financial market APIs consistently place it in the tick-data tier rather than the convenience tier, which is the right read.

Decision rule: choose the polygon api when you need raw, high-resolution history you control. Choose a broker feed when you mainly need a price to act on. If you are still picking a stack, our breakdown of choosing a broker with an API covers that fork in more detail.

How much does the Massive API cost now?

Stock plans cost $0, $29, $79, or $199 per month billed monthly, with annual billing saving 20 percent, per Massive pricing in October 2026. Options, indices, currencies, and futures are sold as separate subscriptions, so a multi-asset bot adds up fast.

Stocks plan	History	Data timing	Includes	Monthly price
Basic	2 years	End of day	Minute aggregates, reference data, corporate actions, indicators	$0
Starter	5 years	15-minute delayed	Unlimited calls, second aggregates, snapshots, WebSockets, flat files	$29
Developer	10 years	15-minute delayed	Everything in Starter plus trades	$79
Advanced	20+ years	Real time	Trades, quotes, financials and ratios, flat files, WebSockets	$199

Two things worth staring at. First, the jump from Developer to Advanced is $120 a month, and it buys the two things day traders care about most: real-time timing and quotes. Second, individual plans are for non-professionals only. If you trade through an entity, redistribute, or build a product on top, you are in business pricing, which is quoted separately.

Common mistake: budgeting $29 for massive api pricing and discovering your options scanner needs its own subscription. The stock plan covers stocks. That is all it covers.

For the full current numbers, read the Massive pricing page directly before you commit, because polygon io pricing has changed more than once since launch and any figure in a blog post has a shelf life.

What do you get on the free plan versus paid plans?

The free Stocks Basic plan is a real free tier, not a 14-day trial, but the 5 calls per minute limit makes it a learning tool rather than a production feed. Paid plans replace that meter with unlimited API calls and add the delayed or real-time timing, WebSockets, snapshots, and flat files that an actual bot needs.

What the free plan includes, per Massive pricing:

• Every U.S. stock ticker
• 2 years of historical data
• End-of-day data and minute aggregates
• Reference data, corporate actions, technical indicators

What the free plan does not include: WebSockets, snapshots, trades, quotes, and flat files. So you can backtest a daily swing strategy, and you cannot run an intraday scanner. That is a clean, honest line.

Rate limits and quotas, plainly stated: Basic is metered at 5 requests per minute. Starter, Developer, and Advanced are described as unlimited API calls on the Massive stocks page, which means there is no published requests-per-minute number to design around. Unlimited access to the endpoint is not an unlimited license to republish the data, and people confuse those two constantly.

Edge case: 5 calls per minute does not mean 5 symbols per minute if you use grouped daily endpoints. One request can return the whole market’s end-of-day bars. Batch your reads and the free tier stretches further than the number suggests.

Is the polygon api real time or delayed?

Delayed on Starter and Developer, real time on Advanced. The 15-minute delay on the $29 and $79 plans is the single most misunderstood fact about current polygon api pricing, confirmed on the Massive pricing page.

Why it matters, depending on what you are building:

• Backtesting and research: delayed is irrelevant. Historical bars are historical bars.
• Swing and position trading: delayed is usually fine. A 15-minute lag on a multi-day hold is noise.
• Day trading, scalping, options flow: delayed data is a history lesson. You cannot read the tape 15 minutes late and call it an entry.

If you are building anything that reacts to a breakout or a gap fill in real time, the $199 tier is the floor, not an upgrade. Anyone selling you a scalping bot on delayed data is selling you a story. Our piece on what the real results show for automated trading bots gets into why latency assumptions quietly destroy backtests.

Decision rule: if your strategy’s average hold is under one hour, pay for real time or do not trade it live.

What changed recently with the polygon api?

Three changes matter in 2026: a remote MCP server for AI assistants, pay-per-request access for agents, and a coming shift in U.S. market hours that will break sloppy session logic. All three are documented on the Massive blog.

• September 15, 2026: remote MCP server. AI assistants can access U.S. market data over OAuth without installing software or having you paste an api key into a chat window. That is a security improvement, not just a convenience one, because the most common way retail developers leak credentials is pasting them somewhere public.
• September 1, 2026: x402 payment access with Coinbase. Agents can request stock data per request, paid in USDC through agent.massive.com, with no conventional account. Crypto shows up here only as a payment rail, and it is the only reason it comes up in this article.
• December 6, 2026: the SIPs move to 23 hours a day, five days a week. Massive announced the change in August 2026. Anything that streams, aggregates by session, or checks market status should be retested. If your bot assumes a 9:30 to 4:00 world with a tidy overnight gap, that assumption has an expiration date.

The changelog also shows ordinary maintenance work: an October 2 entry adding MX2 Options and IEX Options exchanges, and a September 22 correction to Tape B listed-security values, both visible on the Massive stocks page. Boring entries like a data correction are a good sign. A vendor that publishes its own mistakes is a vendor you can audit.

Practical takeaway: put the December 6 date in your deployment calendar now and write a test that feeds your bot a 23-hour session. Find the problem in staging, not with real money on the line.

How do you get a polygon api key and use polygon in python?

Create a free account at Massive, confirm the email, and the dashboard issues your polygon api key immediately. You then pass it as an apiKey query parameter or as a bearer token in the Authorization header, and the official Python client wraps both so you never build URLs by hand.

The short version of getting started:

1. Create the account. Free Stocks Basic, no card required.
2. Copy the key from the dashboard. Treat it like a password. Store it in an environment variable, never in a committed file.
3. Pick your access method. REST for pulls, WebSockets for streams, flat files for bulk history.
4. Install the official Python client. It is published under the polygon-io GitHub organization, which still hosts the SDKs after the rebrand.
5. Make one test call. Request daily aggregates for a single ticker, read the status field, and confirm the record count matches what you asked for.
6. Add a retry with backoff. Not optional. Free-tier throttling is the first problem you will hit.

Using polygon in python is deliberately dull. You create a client with your key, call a method like the aggregates method with a ticker name, a date range, and a timespan, then read the list of bars it returns straight into a dataframe. API cataloguers have mapped the surface area in machine-readable form, including an API Evangelist readme for the polygon-io definitions and a parallel massive-com repository, which is a decent way to see every endpoint listed in one place.

If Python is where you are headed anyway, our guide to Python algorithmic trading covers the surrounding scaffolding, and OpenBB can read Massive as a data provider if you would rather not write the plumbing yourself.

Common mistake: hardcoding the key, pushing to a public repo, then wondering why usage spiked. Rotate the key from the dashboard the moment that happens.

What are the most common polygon api problems and how do you fix them?

The most frequent problem is a 429 rate-limit response on the free plan, followed by empty result arrays caused by bad date parameters and entitlement errors from requesting data your plan does not include. All three are configuration problems, not outages.

Problem you see	Likely cause	Fix
429 Too Many Requests	More than 5 calls per minute on Basic	Add backoff, batch with grouped endpoints, or upgrade
Empty results array	Date range lands on a market holiday or future date	Validate the calendar before you read the response
403 or entitlement error	Requesting trades, quotes, or financials your plan excludes	Check the plan’s feature list, not the docs’ endpoint list
Data looks 15 minutes stale	Starter or Developer plan, by design	Upgrade to Advanced for real time
WebSocket connects then drops	Too many concurrent connections on one key	Fan out from one subscriber process
Field absent from the response	API omits empty fields instead of returning null	Use safe key access in your parser

A fourth problem is architectural rather than technical: polling an endpoint in a loop when you should be streaming, or streaming when a nightly flat file download would do. Every unnecessary request is a problem you created for yourself.

Edge case: after December 6, 2026, a “market is closed” check that reads the clock instead of the market status endpoint becomes a live bug. Read the status, do not assume it.

How does Massive compare to Alpaca, Databento and Finnhub?

Massive is the specialist tick-data vendor, Alpaca bundles data with brokerage, Databento bills by usage, and Finnhub competes on a generous free tier. Pick by what you are actually short of: money, latency, or history depth.

Provider	Best for	Real-time path	Trade-off	Entry price
Massive (formerly Polygon.io)	Deep tick history, flat files, multi-asset	Advanced plan	Real time gated behind the top individual tier	$0 free, $199 real time
Alpaca	Builders who want data and execution in one account	Free IEX feed, or full SIP	Free feed is one exchange, not the full market	$0 free IEX, $99 SIP
Databento	Teams with spiky, precise data needs	Usage based	Cost is variable, so budgeting is harder	Usage based
Finnhub	Beginners and prototypes	Paid tiers	Thinner tick-level depth than a dedicated vendor	Free tier available

Comparison writeups such as WalletInvestor’s look at Polygon.io against Twelve Data and Alpha Vantage land in a similar place: the polygon api wins on raw data quality and loses on convenience features that cheaper services throw in.

Decision rule: if you already trade through Alpaca, start with its free IEX data and only buy a second feed when a real limitation blocks you. Our Alpaca API guide for beginners walks that path, and the Barchart API breakdown covers a third option if fundamentals and futures matter more than ticks.

Two feeds for the same asset class is the most common waste in a retail stack. Pick one, learn it properly, then add.

Top 5 favorite features

The five things that keep developers on Massive after the rebrand are flat files, second aggregates, the snapshot endpoint, the published changelog, and the new agent access methods.

1. Flat files. Bulk historical downloads instead of 40,000 paginated requests. This single feature saves days of backfill work.
2. Second aggregates. Available from Starter up, which is unusually generous for a $29 plan.
3. Snapshots. One request returns the current state of a long list of tickers, which is how you build a watchlist refresh without melting your rate limit.
4. An honest changelog. Exchange additions and data corrections published with dates on the Massive stocks page. Transparency you can verify beats transparency you are promised.
5. MCP and x402 access. As of September 2026, an AI assistant can read market data over OAuth or pay per request, per the Massive blog. That is a real answer to the “my assistant needs data but I will not paste my key” problem.

What we like and what we don’t like

The data quality and history depth are the strongest parts of the product. The pricing structure and the per-asset-class billing are the weakest, and real-time data sitting behind a $199 tier prices out most small accounts.

What we like:

• Backward compatibility after the rebrand, so existing integrations kept working
• A genuinely useful free tier for research, not a trial
• Flat files and deep history, 20-plus years on Advanced
• Unlimited API calls on paid individual plans, which removes rate-limit math

What we don’t like, and these are real drawbacks:

1. Real time starts at $199. The $29 and $79 plans are delayed, which makes them useless for intraday execution. For a trader with a $5,000 account, that subscription is a serious percentage of the account.
2. Each asset class is a separate bill. Stocks, options, indices, currencies, and futures are priced independently per the pricing page, so a multi-asset bot can cost several hundred dollars a month before you place a trade.
3. The rebrand created a documentation seam. Docs, SDKs, and domains are split across the old and new names, and the SDKs still live under the polygon-io organization. Workable, but confusing for a first-time reader.
4. No execution layer. Data only. You still need a broker, which means a second integration and a second set of credentials.

What do real users say about the polygon api?

We did not find user quotes we could verify with a permalink, so none appear here. What can be checked is third-party coverage and the vendor’s own public record, and both point in the same direction: strong data, premium pricing.

Independent API directories continue to list the service under both names, including apis.io’s Massive provider entry and the itechguides product listing for Massive, formerly Polygon.io. Developer roundups such as apilayer’s financial market API comparison place it among the serious options for real-time data rather than the budget tier.

Searches for polygon api reddit and polygon api github mostly lead to the same two conversations: people asking whether delayed data is good enough, and people asking which SDK version to pin after the rename. Both are answered above. For how we read community sentiment in general, see our approach to honest AI stock tool reviews and what real users say.

Position: thin public discussion is not proof of quality in either direction. Verify with your own test calls on the free tier before you pay.

Who should use the polygon api, and who shouldn’t?

Use it if you are building a data-hungry system and you need tick-level history you control. Skip it if you need a price to act on today and you already have a broker who gives you one.

Good fit:

• Backtesters who need 10 to 20 years of minute bars in bulk
• Options flow and microstructure researchers who need quotes, not just bars
• Multi-asset builders who will pay per asset class for consistent formatting
• Developers building AI agents that need authenticated market data access

Poor fit:

• Beginners who want a chart and a scanner. Use a platform, not an API.
• Intraday traders on a small account who cannot justify $199 a month
• Anyone who wants data and execution from one vendor
• Anyone planning to redistribute the data, which individual plans do not permit

If you are earlier in the process, our pieces on what retail traders get wrong about algorithmic trading AI and what to know before you automate trades will save you more money than any data plan.

Not the polygon api you were looking for?

Three different products answer to the name Polygon, and only one of them is a market data service. If you searched polygon api looking for blockchain nodes or a competitive programming problem archive, the Massive docs will not help you.

• Polygon PoS, the Ethereum scaling network. The polygon pos chain and polygon mainnet are accessed through RPC providers, not through Massive. A wallet api or token api for POL balances is a blockchain service api, a completely separate thing from a stock data feed. Anyone asking whether the token will hit a price target is asking about this Polygon.
• Polygon on Codeforces, the problem preparation system. It has nothing to do with either, and its own API is built around problem authoring rather than market data.

Different Polygon, different problem, different API. If your goal is stock data for a trading bot, the polygon api from Massive is the one you want.

Our Take

Massive is still the best raw U.S. market data API most retail developers can buy, and the $199 real-time gate makes it the wrong first purchase for most of them. Start on the free tier, prove your strategy on end-of-day and delayed data, and pay for real time only when latency is the actual constraint.

The rebrand itself was handled well. Keeping api.polygon.io alive so existing integrations kept working was the correct call, and the published changelog with its data corrections earns more trust than a glossy announcement ever could.

What should worry you is structural, not cosmetic. Five separate asset-class subscriptions, a real-time tier that costs more than most small accounts can defend, and a December 6, 2026 market-hours change that will quietly break session logic in a lot of hobby bots. None of that is a reason to avoid the service. All of it is a reason to read the pricing page yourself instead of trusting a number in a blog post.

Systems over hacks. A cheap feed with a written process beats an expensive feed with no stop loss and no position sizing rules. The data is the input, not the edge.

Before you buy: run 30 days of paper trades on free end-of-day data, log every entry and exit, then check whether a 15-minute delay would actually have changed a single decision. If it would not, you just saved $170 a month.

Built something on the polygon api? We catalogue and review the tools retail traders actually use. If you have shipped a bot, a scanner, or a data pipeline worth looking at, submit it for review.

Conclusion

The polygon api did not get worse when it became Massive. It got a new name, a clearer changelog, and some genuinely useful agent access methods, while keeping the old endpoints alive so nothing broke. The pricing is the real story: free for research, $29 and $79 for delayed data, $199 for the real-time feed that intraday strategies actually require.

Three things to do this week:

1. Create a free Stocks Basic key and make one aggregates call for a single ticker.
2. Write down whether your strategy’s average hold time makes a 15-minute delay irrelevant. Be honest.
3. Add a test that simulates a 23-hour trading session before December 6, 2026.

Then stop shopping for feeds and go write your rules.

Related reading

• Alpaca API for beginners
• Barchart API breakdown
• Python algorithmic trading
• What is options flow

Small print: this company changed its entire name in one afternoon. A plan price can move just as fast.

This article is education, not financial advice. Market data plans and prices change, so verify every figure on the official pricing page before you subscribe.

Your market edge starts with the right tool. Stay alpha.