> For the complete documentation index, see [llms.txt](https://docs.delphi.fyi/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.delphi.fyi/what-is-delphi.md).

# What is Delphi?

Delphi is an information market platform where you can get odds on almost anything you can put into words, and trade on them.

## Delphi, the Platform to Predict Anything

Type what you think will happen. Delphi turns it into a market question, gives you odds, and lets you take the other side.

If you think the odds are wrong, you trade against them. Delphi opens the market, funds it, and settles it automatically when the outcome is known. Each market is settled by AI against rules published before anyone trades.

<div data-with-frame="true"><figure><img src="https://3382892862-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Ft59pTERBDU0JfP4e2nEP%2Fuploads%2FPgKWeuHLeaumMDGA4JKj%2Fimage.png?alt=media&amp;token=36671590-ebca-4f72-a6b5-1ae07f83a775" alt=""><figcaption></figcaption></figure></div>

### Two Ways to Create Markets

There are two ways to create markets in Delphi.

1. **You ask.** Write a prediction and Delphi's forecast prices it. If you disagree with the odds, you trade against them, and Delphi opens the market with you as its first trader. These markets are LMSR markets only. See [Predict Anything](/predict-anything.md).
2. **You build.** Write the question, outcomes and settlement terms yourself. This is the path for custom terms, multi-outcome markets, or predictions that Delphi won't open for you, and it makes you the market's creator, earning a share of its trading fees. You can choose to create an LMSR or DPM market. See [Create a Market](/create-a-market.md).

Either way you end up with the same kind of market: anyone can trade it, and it settles through the same automated process.

### Market Mechanisms

Every Delphi market uses one of two mechanisms: **\[1]** LMSR (recommended) or **\[2]** DPM.

* **LMSR (recommended):** The price of an outcome is its probability. Each winning share pays $1 when the market settles. If it closes with *No Outcome*, shares pay their final market price instead.
* **DPM:** Traders buy into a pool, and winners split it.

Learn more about LMSR and DPM markets [here](https://docs.delphi.fyi/pages/Uty4JlTiuqWmOgRXjXGS#id-3.1-lmsr-and-dpm-market-mechanisms).

### Settled by AI, Verifiable by Anyone

Every market declares how it will be decided before anyone trades: the question, the outcomes, the data sources, and the settlement criteria. They're shown on the market page and can't be changed once the market is live.

When trading closes, Delphi's [automated settlement workflow](/how-ai-settlement-works.md) evaluates the available evidence against those rules and submits the result on-chain. Resolution doesn't rely on a central team or a single opaque oracle. Two frontier models independently review the evidence, and an open-source judge model running inside Gensyn's [Reproducible Execution Environment (REE)](https://docs.gensyn.ai/tech) makes the final decision.

#### Verifiable Settlement

REE produces a reproducible receipt for the judge's computation, recording the model, inputs, output, and a cryptographic hash. Anyone can use the receipt to re-run the computation and confirm the recorded model produced the recorded output from the recorded inputs.

This proves the computation happened as stated. It doesn't prove every source or market definition was correct.

{% hint style="info" %}
Read more about what verification does and does not prove in [How Automated Settlement Works](/how-ai-settlement-works.md).
{% endhint %}

Together with the market's public question, outcomes, data sources, and criteria, the receipt makes settlement independently inspectable.

### How It Works

1. **A market is opened.** Either Delphi opens it when you challenge a forecast on the [Predict Anything](/predict-anything.md) page, or a creator builds it in the create flow.
2. **People trade.** Participants buy and sell positions on the outcomes they believe are most likely, and prices move with every trade.
3. **Trading closes.** It closes at a set time, before the outcome is knowable.
4. **Delphi settles the market.** The analyst and judge workflow reviews the declared rules and the available evidence.
5. **The result goes on-chain.** The market page shows the final outcome, payouts, and settlement reasoning.
6. **Winners claim their payouts.** If no winning outcome is declared, the market closes with *No outcome* and positions are paid at their final market prices.

### Where Delphi Runs

Delphi's markets are information markets on Gensyn mainnet, traded in USDC. Anyone can get odds and view markets. Trading and market creation aren't available in certain locations, including the United States and its territories. You can find the full geoblocked region list in [Help](/help.md).

### What's Next

* Visit [Get Started](/get-started.md) to sign in and add funds.
* Make your first prediction on [Predict Anything](/predict-anything.md).
* Browse and trade existing markets on [Explore & Trade](/explore-and-trade.md).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.delphi.fyi/what-is-delphi.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
