Foundational layer for vol research

Institutional-grade
volatility infrastructure.
For the agentic age.

SPX · 13 Mar 2020 · 20 Mar expiry · 09:32–15:57 ET · 32-second replayDesigned for agentic AI research workflows.

OptionSpace / Python research toolkitRecorded examples · SPY · 10 Apr 2025

study_spy.pyPython · optionspaceRecorded session

import optionspace as osp
from optionspace import charts
client = osp.Client()

› smile = client.smile(
    "SPY", "2025-04-10",
    expiration="2025-04-17",
    snapshot="15:57",
)
charts.smile_chart(
    smile, theme="light"
).show()

Fetch the SPY 17 April 2025 expiry at 15:57 on 10 April, then plot the fitted smile with the OptionSpace Python package.

Result details

766 option rows · 322 curve points
historical fitted observations

Out [1] · Figure

Fitted smile · 15:57 ET

Enlarge ↗

Actual OptionSpace output: SPY fitted smile with quote IV bands, 10 April 2025, expiry 17 April

766option observations

54.78%ATM implied volatility

7 DTE17 Apr 2025 expiry

The fitted curve and quote bands come from the same historical snapshot. Inspect the shape, the spread and the extrapolated wings.

Actual package commands and recorded study outputs. No live API calls from this page.Explore more research examples

One pricing engineFitted data & GreeksPlatform for research & idea generationAccessible via API, Python & MCP

Why OptionSpace

Accurate measurement is hard.
That’s what makes it an edge.

Pratik Mathur

Founder & CEO, OptionSpace
Former Global Head of Options at Virtu Financial

OptionSpace is built with a mission to bring market maker grade volatility pricing & tooling to the wider market.

It starts with our fitting-pricing engine. Implied borrows, SPX implied rates, early exercises, discrete dividends, proper settlement lag accounting & vol time decay.

Getting the details right is not an afterthought, it forms the basis of everything we do.

The Vol Engine

Inspect the fitted curve against observed quote bands: in the wings, around earnings, through market stress and minute by minute.

Wings

Arbitrage free wings

Beyond the quoted book, tails extend under Lee-moment discipline, arbitrage free by construction.

SPX and SPY smiles in normalized-strike space. The fitted curve is solid where options are quoted and dashed beyond, continuing smoothly into both tails. — Fig. 03SPX and SPY, expiry 30 January 2026, 10 November 2025 at 15:57 ET. Solid where the market quotes, dashed where it does not.

Shape

AMZN 1 DTE, 26 April 2018. The session before earnings: the event smile changes through the day. Replay shows every third accepted snapshot and the close.

Build a session movie in Python

The displayed movie uses the Python toolkit with captured ClickHouse observations. Fetch the available snapshots, then render the same session.

import optionspace as osp
from optionspace.movies import DayMovie

client = osp.Client()
atm = client.atm_intraday(
    "AMZN", "2018-04-26", expiration="2018-04-27"
)
smiles = [client.smile(
    "AMZN", "2018-04-26", expiration="2018-04-27", snapshot=t
) for t in atm["snapshot_time"]]
movie = DayMovie(smiles, stride=3, include_last=True, xlim=(1250, 1750))
movie.render_frames("shape-frames/")
movie.encode("shape-frames/", "shape")

Stress

SPY and SPXW smiles on 13 March 2020. Bid/ask bars are tens of vol points wide and a third of the SPXW chain has no bid, yet one curve fits both books. — Fig. 0413 March 2020 at the close, the March monthly at 7 DTE. Pale bars are quotes with no bid at all (102 of 322 on SPXW) where the ask is a ceiling and nothing is a floor.

Speed

SPY 0 DTE, 9 April 2025, 12:00–14:30 ET. 150 accepted one-minute fits. The 13:19 fit is excluded by the data-quality checks. The time axis compresses excluded snapshots; labels show actual ET. The curve axes stay fixed throughout the replay.

Build a session movie in Python

The displayed movie uses the Python toolkit with captured ClickHouse observations. Fetch the available snapshots, then render the same session.

import optionspace as osp
from optionspace.movies import DayMovie

client = osp.Client()
atm = client.atm_intraday(
    "SPY", "2025-04-09", expiration="2025-04-09"
)
atm = atm[atm["snapshot_time"].between("12:00", "14:30")]
smiles = [client.smile(
    "SPY", "2025-04-09", expiration="2025-04-09", snapshot=t
) for t in atm["snapshot_time"]]
movie = DayMovie(smiles, atm=atm, xlim=("z", -5, 3.5))
movie.render_frames("speed-frames/")
movie.encode("speed-frames/", "speed")

The same data, through your agent

Start with a question.
Follow it into the data.

Compare dates, inspect a curve, or put an exposure in historical context. OptionSpace brings the data and research tools into your own workflow.

Research, in your own words.Claude Code + OptionSpace MCP · real recorded tool calls

Chart returned by the recorded skew comparison

Compare skew across names and dates.

One ordinary question compares SPY and NVDA for the same listed expiry, using the production skew method and showing the previous-session change.

SPY +4.66 · NVDA +9.76 vol pts

10 Apr versus 9 Apr 2025 · expiry 17 Apr. Resolved snapshots accompany the comparison.

Recorded Opus 5 responses, rendered as text. Waiting shortened; setup and intermediate tool output omitted. Select a study to replay. Charts stay available below.Explore more research tools →

One engine. Two products.

Take the data into your workflow.
Or explore it in ours.

01 / Data

Build on the numbers.

Fitted option smiles and Greeks, constant-maturity metrics, and implied rates for quantitative research. In a high performance managed DB.

Use it for: Historical analysis, model inputs and your own systematic studies.
Work with: REST API, Python DataFrames, local MCP and a data inspection portal.

Explore the data

02 / Platform

Work through the question.

Inspect the surface, place volatility in historical context, and investigate an idea with visual research tools.

Use it for: Surface inspection, event studies, screening, historical simulations and idea generation.
Work with: An interactive research environment, built around the same engine.

Explore the platform

Stay in the loop

Sign up to get updates
about our launch.

Or reach us at [email protected] for a conversation about your needs.