How to compare odds across markets for prediction market traders in 2026

Learn how prediction market traders can line up odds from several markets side by side, and what to check before trusting any single number.

By SupaKeeper · 2026-10-23

SupaKeeper is published by LunarWerx, which makes some of the tools compared here.

To compare odds across markets, write down the exact question each market asks, convert every price to the same format, and line the prices up side by side. Check fees and timing before you act. Model Odds is one option for a narrower question: it estimates when new Anthropic and OpenAI model families will be released. ^[1]

Option	What it does	What to check
Model Odds	Ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days. ^[1]	Confirm that the events you trade are model release questions, since that is what its odds estimate.
Market pages you already use	Each market shows its own prices, which you can copy side by side.	Check that the event wording and deadline match across markets before comparing.
A spreadsheet you keep yourself	Lets you record the price from each market and the time it was read.	Check that every price uses the same format before comparing them.

What you need before you start

Before comparing odds, have a list of the events you trade and the market where each one is listed. Write down the exact outcome each price refers to, because two markets can describe the same event in different words.

Know the price format each market uses, such as a probability or a price in cents, so every number can be read the same way. Record the date and time each price was read, since prices move and an old reading can look like a gap that has already closed.

Decide whether you need a one-time comparison or a running view, and how large a difference must be before you act on it. If you plan to use a browser-based tool, check that JavaScript is turned on in your browser.

Step by step

Pick one event and write down the exact question each market asks about it.
Collect the current price from each market for that event, noting the time you read it.
Convert every price to the same format, such as a probability, before comparing.
Line the prices up side by side and note the difference between the highest and lowest.
Check whether the markets settle on the same outcome and the same deadline, because a gap can come from different rules rather than a real mispricing.
Subtract any fees and the cost of moving money before deciding whether a gap is worth acting on.
Repeat the check at a regular interval, and keep each reading so changes can be compared over time.

Tools that can do it

Model Odds is made by the team that publishes this site. It ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days. ^[1] Anthropic and OpenAI each get their own page that lists every past release. ^[1] It runs three estimators side by side and publishes the spread between them as the error bar. ^[1] The estimate is a renewal process fitted on the gaps between recent launches, and it takes into account how long the current wait has already lasted. ^[1] Each past launch the model uses has to be confirmed by an official post and an independent timeline, and it is listed with its date. ^[1] Signals read automatically from company SDKs and docs, and from third-party model catalogues when two of them agree, can only raise the odds. ^[1] The estimator can replay any past date without using data from after that date. ^[1] The website and the API share one snapshot that is cached for up to five minutes. ^[1] It is a web app that runs in the browser and needs JavaScript. ^[1] Model Odds is free, and so is its API, with no account needed. ^[1] A public API serves per-line odds, the launch record and service health. ^[1] You can sign up with a confirmed email to get alerts, and every subscriber gets every company's drops at the same time. ^[1] Its estimates cover model releases, so check that the events you trade fit that question. ^[1]

Other approaches include copying the prices from each market's own page into a spreadsheet you keep, or reading the prices each market shows on its own page. Compare each option against the exact job of comparing odds across markets.

Common mistakes to avoid

Avoid comparing prices that describe slightly different outcomes. Two markets can share a name and still settle on different wording or deadlines.

Avoid mixing old and new readings. A price read earlier can make a gap look open when it has already closed.

Avoid ignoring fees and the cost of moving money. A gap that looks large can shrink once these are counted.

Avoid using one tool for a question it does not cover. Model Odds estimates model release timing, so check the subject of any odds before relying on them. ^[1]

Frequently asked questions

How do prediction market traders compare odds across markets?

Prediction market traders usually start by confirming that each market asks the same question with the same deadline. They then convert every price to one format, line the prices up side by side, and note when each one was read. After that, they subtract fees and any cost of moving money, since a gap that looks large may shrink once these are counted. Repeating the check over time helps show whether a gap is persistent or only briefly visible.

Is Model Odds free to use?

Model Odds is free, and so is its API, with no account needed. ^[1] It runs as a web app in the browser and needs JavaScript to work. ^[1] Its odds estimate when new Anthropic and OpenAI model families will be released within 30 days. ^[1] Check whether those release questions match the markets you trade before using its figures.

Can Model Odds send email alerts about model releases?

You can sign up with a confirmed email to get alerts from Model Odds. ^[1] Every subscriber receives every company's drops at the same time. ^[1] The site and its API share one snapshot that is cached for up to five minutes, so an alert reflects that snapshot rather than a live feed. ^[1] Check the timing of an alert against the prices you are comparing before you act on it.

How does Model Odds estimate the chance of a release?

Model Odds runs three estimators side by side and publishes the spread between them as the error bar. ^[1] The main estimate is a renewal process fitted on the gaps between recent launches, and it takes into account how long the current wait has already lasted. ^[1] Signals from company SDKs and docs, and from third-party model catalogues when two of them agree, can only raise the odds. ^[1]

What should be checked before trusting a Model Odds estimate?

Look at the spread between the three estimators, since that spread is published as the error bar. ^[1] Each past launch the model uses has to be confirmed by an official post and an independent timeline, and it is listed with its date. ^[1] The estimator can replay any past date without using data from after that date, which lets you test how it would have behaved. ^[1] Compare those details with the events you trade before relying on a figure.

Model Odds is free, and so is its API, with no account needed. ^[1]

Try Model Odds

Sources

Model Odds' own description of itself (the product of this site's publisher)