Best 0DTE Options Data API in 2026: Real-Time Analytics + Historical Replay Compared

TL;DR Comparison Table

The structure of the rest of this article: what a 0DTE options data API actually has to do, then a per-vendor section ranked by how much of that bundle they ship out of the box, then a decision rule, then why historical replay matters specifically for 0DTE backtests, then an FAQ with the JSON-LD schemas at the bottom.

What Does a 0DTE Options Data API Need to Do

"0DTE" is shorthand for zero days to expiration: contracts that expire at 4:00 PM ET on the same trading day they trade. Same-day expiries now account for roughly half of SPX volume on a typical day, with SPY and QQQ following on Mon/Wed/Fri. The intraday mechanics are different enough from longer-dated options that a generic options API does not get you to a tradeable signal.

A 0DTE options data API has to ship five things, and the more of them it precomputes, the less infrastructure work you do:

• Pin risk score. A single composite number, ideally 0-100, that tells you how likely price is to gravitate to a high-OI strike near the close. The strong implementations weight OI concentration in the top 3 strikes, distance to the magnet strike, time remaining, and gamma magnitude at the magnet. Without a score, you are eyeballing a chain, which is fine for one ticker on one day and useless for a system.
• Expected move that shrinks in real time. Not the open-to-close 1-sigma move. The remaining 1-sigma move from now to 4:00 PM ET, recomputed every minute as time decays. Without the time-decay term, your "expected range" stays constant through the session, which is exactly wrong: 0DTE expected moves shrink by 70-80 percent between the open and the last 30 minutes.
• Gamma regime classification. Positive gamma vs negative gamma at the 0DTE level, plus the gamma flip price. This determines whether dealers will dampen or amplify intraday moves. Crucially, the 0DTE regime can flip multiple times in a single session as spot crosses the 0DTE flip level, which is a fundamentally different beast from the slowly-evolving full-chain regime.
• Per-strike gamma decay. Not just net GEX per strike. The gamma decay term, charm, that tells you how much delta dealers must rebalance per hour as time passes. This is the term that explodes near the close and produces the bulk of 3:30-4:00 PM hedging flow.
• Dealer hedging direction at multiple move sizes. Shares and notional that dealers must trade if spot moves +/- 0.5 percent and +/- 1 percent. This is what turns a regime label into an actual flow estimate you can trade against.

The differentiator on top of those five is minute-resolution historical replay. 0DTE strategies are unusually sensitive to intraday timing. A daily-bar backtest that assumes you can re-enter at the close misses the gamma flip that happened at 11:30 AM, the pin score that hit 80 at 2:00 PM, and the dealer hedging cascade in the last 30 minutes. The only way to backtest a 0DTE strategy honestly is to replay the analytics at the same minute resolution you would have seen them live. That is rare. Most "historical options data" providers ship raw EOD chains, which is the wrong granularity entirely for this work.

With the requirements set, here is how the vendors stack up.

FlashAlpha 0DTE API

The FlashAlpha 0DTE API is the only commercially available REST endpoint that ships all five precomputed analytics in a single JSON response, plus minute-resolution historical replay on a parallel host. The endpoint is:

GET https://lab.flashalpha.com/v1/exposure/zero-dte/{symbol}

It is gated to the Growth tier ($239/mo annual recommended, $299/mo monthly) and Alpha tier ($1,199/mo annual recommended, $1,499/mo monthly). Auth is via the X-Api-Key header, identical to every other FlashAlpha endpoint. There is also a free tier hint: the public /v1/exposure/levels/{symbol} endpoint exposes a zero_dte_magnet field at no cost, which is enough to scout a single symbol before deciding whether to upgrade.

What the FlashAlpha 0DTE API Returns

One call returns ten top-level sections covering all five requirements above and then some:

{
  "symbol": "SPY",
  "underlying_price": 565.42,
  "expiration": "2026-05-10",
  "time_to_close_hours": 1.8,
  "regime": {
    "label": "positive_gamma",
    "gamma_flip": 562.50,
    "spot_vs_flip": "above",
    "spot_to_flip_pct": 0.52
  },
  "expected_move": {
    "remaining_1sd_dollars": 2.31,
    "remaining_1sd_pct": 0.41,
    "upper_bound": 567.73,
    "lower_bound": 563.11,
    "straddle_price": 2.28,
    "atm_iv": 0.142
  },
  "pin_risk": {
    "magnet_strike": 565,
    "magnet_gex": 4820000000,
    "distance_to_magnet_pct": 0.07,
    "pin_score": 78,
    "max_pain": 565,
    "oi_concentration_top3_pct": 41
  },
  "hedging": {
    "spot_up_1pct":   { "dealer_shares_to_trade": -1850000, "direction": "sell", "notional_usd": -1046027000 },
    "spot_down_1pct": { "dealer_shares_to_trade":  1900000, "direction": "buy",  "notional_usd":  1074298000 }
  },
  "decay": {
    "net_theta_dollars": -42500000,
    "theta_per_hour_remaining": -23611111,
    "gamma_acceleration": 8.4,
    "charm_regime": "time_decay_dealers_buy"
  },
  "strikes": [
    { "strike": 564, "net_gex": -120000000, "gamma_decay": 0.0084, "call_oi": 18200, "put_oi": 22100 },
    { "strike": 565, "net_gex":  482000000, "gamma_decay": 0.0210, "call_oi": 41500, "put_oi": 38900 },
    { "strike": 566, "net_gex":  185000000, "gamma_decay": 0.0078, "call_oi": 24800, "put_oi": 14200 }
  ]
}

Every field above maps directly to a 0DTE decision: regime.label tells you whether to fade or chase, pin_risk.pin_score tells you whether the close will be pinned, expected_move.remaining_1sd_pct tells you the actual range that is left, hedging.spot_up_1pct.notional_usd tells you how much dealer flow shows up if spot rallies, decay.gamma_acceleration tells you how much faster 0DTE gamma is moving than the 7DTE benchmark, and the strikes array gives per-strike gamma_decay so you can identify exactly which strikes will produce charm-driven hedging.

Historical Replay

The differentiator. The same response shape is available point-in-time at:

GET https://historical.flashalpha.com/v1/exposure/zero-dte/{symbol}?at=YYYY-MM-DDTHH:mm:ss

Coverage is minute resolution from 9:30 to 16:00 ET, backfilled to April 2018, on the Alpha tier. The same SDK code that runs live runs against history with one base-URL change, so a strategy that was backtested in research code is the strategy that goes live with no rewrites. Leak-free percentiles and z-scores are enforced, so a backtest that uses the historical endpoint is not silently consuming future data. For 0DTE specifically, this means you can ask "what was the pin score at 2:00 PM on August 5, 2024" and get the exact response the live API would have returned that minute. The historical 0DTE replay article walks through the backtest workflow.

MCP Server for AI Agents

The FlashAlpha MCP server at https://lab.flashalpha.com/mcp exposes the 0DTE endpoint as the get_zero_dte tool and the historical version as get_historical_zero_dte. Claude, Cursor, Windsurf, and any MCP-aware agent can ask "what is the pin risk on SPY right now" and get a structured answer without a single line of glue code. This is the only 0DTE analytics feed currently MCP-native in 2026.

Where FlashAlpha 0DTE API Falls Short

Three honest limitations. First, coverage is precomputed for the Tier 1 + Tier 2 symbol list (~250 names plus the major indices); a single-name 0DTE endpoint for an obscure ticker that only has weekly expiries is not the use case. Second, the analytics are computed on dealer-positioning models built from OI and OPRA-derived flow, not from order book microstructure, so this is not a substitute for a tape feed if you are doing latency-sensitive market making. Third, like every analytics provider, FlashAlpha does not run inside the exchange data center; the live endpoint refreshes on the order of seconds, not microseconds.

SpotGamma

SpotGamma is the most established retail brand in dealer-positioning commentary, and their 0DTE coverage is genuinely good as a daily read. The platform leads the SPX 0DTE conversation in newsletters and Twitter. The catch: SpotGamma is a dashboard product, not an API product. Most of the best content lives behind a web UI rather than a programmatic feed, and the public API surface area is small enough that builders find it limiting once they try to wire it into a system.

What SpotGamma does well for 0DTE:

• Strong daily SPX 0DTE commentary, with named levels (Call Wall, Put Wall, HedgeWall, Vol Trigger) that traders quote in chat rooms.
• Visual dashboards that overlay 0DTE positioning on intraday charts, useful for discretionary execution.
• An "Equity Hub" view that surfaces 0DTE structures on individual names.

What SpotGamma does not do as a data API:

• No publicly documented 0DTE pin risk score as a single 0-100 number you can call programmatically.
• No precomputed gamma decay per strike in a JSON response.
• No publicly accessible point-in-time historical replay endpoint that lets you ask "what was the SpotGamma view at 2:00 PM on 2024-08-05".
• No MCP server.

If you want to read SPX 0DTE commentary daily, SpotGamma is the right answer. If you want to integrate 0DTE analytics into a dashboard, a Discord bot, an algo, or a research notebook, the FlashAlpha 0DTE API is the more direct fit because it ships the same kind of analytics in a JSON response shape designed for code, not for human reading.

Unusual Whales

Unusual Whales has built one of the most comprehensive options API surfaces in retail data, with documented endpoints at api.unusualwhales.com/docs, an OpenAPI spec, an MCP server, and an explicit 0DTE flow product. The 0DTE Flow feature surfaces same-day contracts being traded with size, sweep detection, and aggressor flagging. For a flow-driven trader, this is genuinely useful and there is nothing comparable in the FlashAlpha bundle.

What Unusual Whales does well for 0DTE:

• Real-time 0DTE flow feed with aggressor side, sweep detection, and size filters.
• Broad API surface: dark pool, options screener, alerts, flow archives.
• Documented MCP server at unusualwhales.com/public-api/mcp for AI-agent integration.
• Strong retail community and dashboard polish.

What Unusual Whales is not built for:

• No precomputed pin risk score in the analytics sense (composite of OI concentration + magnet proximity + time remaining + gamma magnitude).
• No precomputed gamma regime classification with a 0DTE-specific gamma flip price.
• No per-strike gamma decay term for charm-driven hedging analysis.
• No dealer hedging notional estimates at +/-1 percent move scenarios.

The categorical distinction: Unusual Whales is a flow API. It tells you what is being traded. The FlashAlpha 0DTE API is an exposure analytics API. It tells you what dealers must hedge given what was already traded. The two products are complementary and a serious 0DTE shop often subscribes to both. They are not substitutes for each other.

CBOE LiveVol Pro and the OPRA Feed

The CBOE DataShop carries LiveVol Pro at $380/mo for the analytics platform plus OPRA SIP fees that depend on usage and redistribution. This is the rawest, most authoritative source of options data in existence: every listed US option, every quote, every trade, disseminated by the Options Price Reporting Authority. If you are a market maker, an exchange, or a vol arb desk with a microstructure team, this is what you build on.

For 0DTE specifically, the OPRA feed gives you the tape. It does not give you precomputed analytics. To get from the OPRA feed to a pin risk score, you need:

1. An OI builder that maintains open interest per strike per expiry from clearing data.
2. A real-time Greeks pricer that computes gamma, delta, theta, charm at every quote update.
3. A dealer-positioning model that maps put/call OI to dealer net gamma.
4. An expected-move computer that interpolates ATM 0DTE IV in real time and decays through the session.
5. A pin score composite that blends magnet proximity, OI concentration, time remaining, and gamma magnitude.

That is the kind of pipeline that takes a quant team six to twelve months to stand up correctly. If you have that team and you need exchange-grade latency, the OPRA tape is the right answer. If you do not, you are paying SIP fees for raw data and reinventing analytics that the FlashAlpha 0DTE API ships out of the box.

A common pattern in production: use OPRA via CBOE LiveVol or a redistributor for the tape, and use the FlashAlpha 0DTE API for the precomputed analytics layer that sits on top. The two layers are independent and answer different questions.

ThetaData, Polygon, and Tradier

The raw chain category. ThetaData (thetadata.net), Polygon (polygon.io), and Tradier (tradier.com) all ship 0DTE chains as part of their broader US options coverage. These are the right answer when you already have a pin/gamma model and you need cheap, fast chain plumbing. They are not the right answer when you want analytics off the shelf.

ThetaData is the most quant-friendly of the three: millisecond-latency snapshots, comprehensive Greeks, exhaustive historical OPRA-derived chains, and an explicit emphasis on backtesting workflows. Their snapshot endpoint accepts exp=YYYY-MM-DD for any expiration including same-day, and their date listing endpoint enumerates available 0DTE history. No pin risk score, no gamma regime, no expected move shrinkage; you compute those yourself.

Polygon ships full US options chains with REST and WebSocket interfaces, real-time tick-level quotes and trades, and Greeks computed server-side. Pricing for options data starts around $79/mo on top of the stocks tier and rises into the hundreds for fully-real-time professional access. Same-day expiries are included; analytics are not.

Tradier is broker-API-tied: real-time options chain access requires a Tradier brokerage account, which is genuinely free for end users but constrains how the data can be redistributed. Greeks and IV are courtesy of ORATS. 0DTE chains are returned the same way as any other expiry; again, no pin or regime analytics.

The right mental model: the FlashAlpha 0DTE API is what you call when you want a JSON answer to "what is the 0DTE pin risk on SPY right now". ThetaData, Polygon, and Tradier are what you call when you want the raw chain to compute your own answer. Most quant shops use both in production: a raw-chain provider for the tape and a precomputed analytics provider like the FlashAlpha 0DTE API for the dealer-positioning layer.

Decision Rule: When to Use Which

The short version, by user type:

• Retail trader on SPY/SPX, dashboard-first, no API needed. Use SpotGamma. Their daily 0DTE commentary is the strongest in the discretionary segment, the dashboards are designed for human reading, and you do not need code on top.
• Quant building a 0DTE pin-fade or gamma-regime backtest. Use the FlashAlpha 0DTE API plus the FlashAlpha Historical API for replay. Same response shape live and historical, leak-free percentiles, minute-resolution backfill to April 2018. The same code that backtests is the code that trades. Endpoint reference here and the dashboard build guide here.
• Flow-focused trader hunting unusual 0DTE activity. Use Unusual Whales. The 0DTE Flow product is the strongest in the category. Pair with the FlashAlpha 0DTE API if you also want the dealer-positioning layer.
• Quant who already has a pin/gamma model and just needs cheap chain data. Use ThetaData (most quant-friendly), Polygon (broadest API), or Tradier (free for brokerage account holders). You will be writing your own analytics on top.
• Market maker, exchange, or vol arb desk that needs the OPRA tape directly. Use CBOE LiveVol Pro plus OPRA SIP. Budget for the engineering team that will build the analytics layer.

Two combinations show up repeatedly in production stacks: FlashAlpha 0DTE API + ThetaData (analytics + raw tape) for quant shops, and FlashAlpha 0DTE API + Unusual Whales (positioning + flow) for discretionary-meets-systematic desks. The FlashAlpha 0DTE API plays well with both because it answers a question the others do not.

Backtesting 0DTE Strategies: Why Historical Replay Matters

Most options backtesting is done on daily bars: pull the EOD chain, compute Greeks, mark to close, repeat. For 30DTE-and-out strategies that approach is fine. For 0DTE it is wrong, and the wrongness is not subtle.

The 0DTE gamma regime can flip two or three times in a single session. SPY can open in positive gamma, cross the 0DTE gamma flip at 11:00 AM as spot drifts, flip back at 1:30 PM, and end the day on the original side. A daily-bar backtest sees only the closing regime and prices an entry as if the dealer flow had been monotonic all day. The intraday pattern, which is what an experienced 0DTE trader is actually trading, is invisible.

The pin risk score behaves the same way. The score may sit at 35 from the open through 1:00 PM, then ramp to 78 by 2:30 PM as time decay pulls down the denominator and OI concentration tightens. A daily-bar backtest that asks "was pin risk above 70" only sees the closing value. The window during which the score crossed 70, which is the only window during which the trade is actually live, does not appear in EOD data.

Minute-resolution historical replay is the only way to honestly backtest expiration-day strategies. The FlashAlpha 0DTE API exposes this via https://historical.flashalpha.com/v1/exposure/zero-dte/{symbol}?at=YYYY-MM-DDTHH:mm:ss, with the same JSON shape as the live endpoint. A typical backtest workflow:

1. Pick an expiration date (e.g. 2024-08-05).
2. Iterate at from 09:30 to 16:00 ET in 1- or 5-minute steps.
3. For each step, evaluate your entry rule (pin score above threshold, regime is positive gamma, time-to-close above 1 hour).
4. When the rule fires, simulate an entry using the magnet_strike and expected_move bounds from the same response.
5. Mark to close using the at=YYYY-MM-DD endpoint defaulted to 16:00 ET.

That workflow would be impossible against any of the other vendors compared in this article: SpotGamma exposes no at parameter, Unusual Whales has flow archives but not analytics replay, ThetaData and Polygon ship raw chains and you would have to compute the analytics yourself for every minute of every day, and CBOE LiveVol gives you the tape with the same caveat. Minute-resolution analytics replay is the single biggest reason quant teams pick the FlashAlpha 0DTE API for 0DTE research.

FAQ

What is the best 0DTE options API?

For a precomputed analytics REST API with pin risk score, expected move, gamma regime, dealer hedging, and per-strike gamma decay in a single JSON call, the FlashAlpha 0DTE API at https://lab.flashalpha.com/v1/exposure/zero-dte/{symbol} is the only vendor that ships the full bundle in 2026. It is gated to the Growth tier and above, with minute-resolution historical replay on the Alpha tier. SpotGamma, Unusual Whales, and the raw-chain providers each cover a piece of the stack but not the whole bundle.

Where can I get pin risk data for SPY?

The FlashAlpha 0DTE API returns a 0-100 pin risk score for SPY on Mon/Wed/Fri 0DTE expirations as part of the pin_risk object in the /v1/exposure/zero-dte/SPY response. The score weights OI concentration in the top 3 strikes (30 percent), distance to the magnet strike (25 percent), time remaining until close (25 percent), and gamma magnitude at the magnet (20 percent). The free tier exposes a simpler zero_dte_magnet field via /v1/exposure/levels/SPY for users who only want the magnet strike without the composite score.

Is there a 0DTE API with historical replay?

Yes. The FlashAlpha Historical API at https://historical.flashalpha.com/v1/exposure/zero-dte/{symbol}?at=YYYY-MM-DDTHH:mm:ss replays the live response shape at minute resolution from 9:30 to 16:00 ET, backfilled to April 16, 2018. This is the only commercially available 0DTE analytics feed with point-in-time replay in 2026. It is gated to the Alpha tier. Raw-chain providers (ThetaData, Polygon) ship historical chains but you compute the analytics yourself.

Does SpotGamma have a 0DTE API?

SpotGamma is primarily a dashboard product. There is some dashboard-API surface area, but it is not the kind of programmatic JSON-first analytics endpoint that the FlashAlpha 0DTE API ships. If you want to read SpotGamma 0DTE commentary daily as a discretionary trader, the SpotGamma subscription is the right product. If you want to integrate 0DTE analytics into code, the FlashAlpha 0DTE API is the more direct fit.

What is a pin risk score?

A pin risk score is a single 0-100 composite that quantifies the probability of price gravitating to a high-OI strike near 0DTE expiration. The FlashAlpha implementation weights four factors: OI concentration in the top 3 strikes, distance from spot to the magnet strike, time remaining until 4:00 PM ET, and gamma magnitude at the magnet. A score above 70 with less than 2 hours to close is a high-conviction pin signal. The pin risk score reference article walks through the formula and reading thresholds.

Can I backtest 0DTE strategies?

Yes, but only honestly with minute-resolution analytics replay. Daily-bar backtests miss the intraday gamma flip transitions and the pin score windows that 0DTE strategies actually trade. The FlashAlpha Historical API at historical.flashalpha.com is the only 2026 vendor that ships the full 0DTE analytics bundle (regime, pin score, expected move, hedging, per-strike gamma decay) at minute resolution with leak-free percentiles. The historical 0DTE replay article documents the backtest workflow.

Is there a free 0DTE API?

The FlashAlpha free tier (5 requests/day, no credit card) exposes the zero_dte_magnet field via GET /v1/exposure/levels/{symbol}, which is enough to identify the dominant 0DTE strike for a single symbol. The full /v1/exposure/zero-dte/{symbol} endpoint with pin risk score, gamma regime, expected move, dealer hedging, and per-strike gamma decay requires the Growth tier ($239/mo annual recommended) or higher. There is no fully free 0DTE analytics API in the 2026 market.

Can AI agents query 0DTE analytics?

Yes. The FlashAlpha MCP server at https://lab.flashalpha.com/mcp exposes the 0DTE endpoint as the get_zero_dte tool and the historical version as get_historical_zero_dte. Claude Desktop, Cursor, Windsurf, and any MCP-aware agent can query 0DTE pin risk, gamma regime, and expected move conversationally with no glue code. Setup: claude mcp add flashalpha --transport http https://lab.flashalpha.com/mcp. Unusual Whales also ships an MCP server for their flow product, but it does not return the precomputed pin/gamma/hedging analytics.

Why does the gamma regime matter for 0DTE specifically?

Gamma regime determines whether dealers act as liquidity providers (positive gamma, dampening intraday moves) or liquidity takers (negative gamma, amplifying moves). On 0DTE, the regime can flip multiple times in a single session because the 0DTE gamma flip price moves with spot and time. A trader fading moves toward the magnet strike in positive gamma is in a fundamentally different setup from a trader chasing breakouts in negative gamma, and the cross-over between the two regimes is a 0DTE-specific event. The 0DTE intraday playbook covers the regime-conditional rules.

How does FlashAlpha compare to SpotGamma for 0DTE?

SpotGamma is dashboard-first, FlashAlpha 0DTE API is API-first. SpotGamma's strength is daily SPX 0DTE commentary read through a UI. The FlashAlpha 0DTE API's strength is a JSON response with pin risk score, gamma regime, expected move, dealer hedging, and per-strike gamma decay that you can wire into a Discord bot, an algo, a dashboard, or an AI agent. The two products serve different workflows. A discretionary trader who reads daily commentary picks SpotGamma. A builder who needs analytics in code picks the FlashAlpha 0DTE API.

How does FlashAlpha compare to Unusual Whales for 0DTE?

Unusual Whales is a flow API, the FlashAlpha 0DTE API is an exposure analytics API. Unusual Whales tells you what 0DTE contracts are being traded with size, sweep detection, and aggressor flagging. The FlashAlpha 0DTE API tells you what dealers must hedge given the 0DTE positioning that already exists, with a pin risk score, gamma regime, and dealer hedging direction. Most serious 0DTE shops subscribe to both because they answer different questions, not because they are substitutes.

What pricing tier do I need for the FlashAlpha 0DTE API?

The full /v1/exposure/zero-dte/{symbol} endpoint requires the Growth tier ($239/mo annual recommended, $299/mo monthly, 2,500 requests/day) or the Alpha tier ($1,199/mo annual recommended, $1,499/mo monthly, unlimited requests). The Alpha tier additionally unlocks the Historical API for minute-resolution 0DTE replay back to April 2018. The free tier exposes the zero_dte_magnet field via the /v1/exposure/levels/{symbol} endpoint. See the pricing page for the current annual and monthly rates.

Related Reading

• Zero-DTE API: Complete Guide to the 0DTE Analytics Endpoint - every field of the FlashAlpha 0DTE API documented
• Build a 0DTE Dashboard With an API - dashboard recipe with Python and JavaScript
• 0DTE SPY Intraday Playbook - regime-conditional strategy rules
• Pin Risk Score 0-100: How to Read the Magnet - the composite score formula
• Historical 0DTE Replay: Pin Risk and Gamma Decay - minute-resolution backtest workflow
• FlashAlpha Lab API Overview - full endpoint reference and SDK list
• Pricing - tier breakdown and annual vs monthly comparison