Product
Live data Historical data Chrome extension
Datasets
Weather Markets
Docs
Quickstart API reference
Company
About Our mission

Clean data for real-world predictions.

Mostly Right turns historical and live public data like weather, economic releases, financial reports, and satellite data into one consistent shape, with clear sources and timestamps.

Chrome extension
live-vs-archive.py
1from mostlyright import weather
2
3live = await weather.live.latest("KNYC")
4historical = weather.observations(
5    "KNYC", "2025-01-15", "2025-01-15",
6    return_type="list",
7)[-1]
training-table.py
1from datetime import date
2from mostlyright import weather
3from mostlyright.markets.catalog import kalshi_nhigh
4
5day = date(2025, 1, 15)
6contract = kalshi_nhigh.resolve("KHIGHNYC", day)
7table = weather.training_table(
8    contract.settlement_station, date(2025, 1, 12), day
9)

Point-in-time research for prediction markets.

Start with a Kalshi or Polymarket contract. Mostly Right resolves its official reference, keeps the cutoff visible, and puts the outcome beside data available by then.

One contract in. One research table out.

01
Resolve the settlement reference

Map the venue code to its official station, unit, and outcome source.

contract → reference
KXHIGHNY KNYC New York · °F NWS CLI
KXHIGHAUS KAUS Austin · °F NWS CLI
KXHIGHMIA KMIA Miami · °F NWS CLI
HIGH-LDN EGLC London · °C NOAA / WU
HIGH-TOK RJTT Tokyo · °C NOAA / WU
HIGH-PAR LFPG Paris · °C NOAA / WU
02
Build the research table

Place historical inputs beside the official outcome and keep the cutoff visible.

Field Type Meaning
local_standard_date local date station-local standard day
station string canonical ICAO station
decision_time_utc ISO datetime research cutoff
daily_summary_temp_high_f number · °F official daily high label
observed_temp_high_f number · °F highest observation by the cutoff
observation_count integer reports included before the cutoff
knowable_at_utc ISO datetime latest input knowledge time
research table 07 fields
local_standard_date local date

station-local standard day

station string

canonical ICAO station

decision_time_utc ISO datetime

research cutoff

daily_summary_temp_high_f number · °F

official daily high label

observed_temp_high_f number · °F

highest observation by the cutoff

observation_count integer

reports included before the cutoff

knowable_at_utc ISO datetime

latest input knowledge time

From satellite data to financial reports.

Explore public datasets for prediction-market research and predictive modeling, with clear sources and timestamps.

Weather

Live observations, forecasts, climate records, and official daily outcomes from public weather sources.

Explore weather →
Markets

Kalshi and Polymarket discovery, market history, contract research, and settlement data.

Explore markets →
Satellite data

Earth observations mapped to weather stations for use as model inputs.

Coming soon
Economic releases

CPI, PPI, payrolls, GDP, jobless claims, and Fed decisions with first-print vintages preserved.

Coming soon
Financial reports

Earnings-call transcripts, point-in-time facts, and prediction-market mention resolution.

Coming soon
Politics

Election results, polling, and policy calendars are next, with source identity and release time kept explicit.

Coming soon

Train on the past. Run on the present.

Historical and live methods return the same core fields and units. Every row states which source produced it.

What every read includes

Rows validate against versioned schemas and include source and timing fields, so you can inspect the exact inputs later.

schema / checks 06 checks
01
Direct-source reads

Public APIs run from your runtime; hosted delivery stays explicit where required.

02
Versioned schemas

Rows validate against versioned JSON Schemas.

03
Local-first cache

Pinned inputs stay reusable for repeatable research across Python and TypeScript.

04
Matching Python and TypeScript APIs

Shared methods use matching snake_case fields and source tags in both SDKs.

05
Source named on every frame

Every frame names the feed that produced it, so you can detect a source change.

06
Temporal safety

Arrival-time checks keep data published after your cutoff out of the backtest.

Browser extension

See live observations, past observations, and forecasts directly on supported Polymarket and Kalshi weather markets.

kalshi.com / markets / KXHIGHNY
Highest temperature in New York today?
Climate and Weather · Daily Temperature

Build your first research table.

Request access and tell us what you are building.

Read the docs
// request access

Request early access

Mostly Right is currently available by request. Tell us a bit about you and what you are building, and we will be in touch.

I'm a

One email from the team. No spam.