# What to use instead of Augur

The contracts are live and the fork ran to completion in August 2026 — but no application has relaunched, and REP that missed the deadline is stranded.

*https://predictionmarkets.tools/alternatives/augur*

## The subject

| Field | Value |
| --- | --- |
| Product | Augur |
| Status | restricted |
| Website | https://www.augur.net/ |
| Category | Prediction Market Venues |

A decentralised prediction-market protocol on Ethereum where disputed outcomes are settled by REP holders staking against each other, and in the last resort by splitting the protocol into one universe per outcome.

## Replacements

*Ordered by how much of the original’s job each one covers, closest first. Paid placement does not affect this order.*

1. [Polymarket](https://predictionmarkets.tools/tools/polymarket.md) — Self-custody on Polygon, no identity check at signup, and the only resolution here a stranger can enter — a bonded proposal, a challenge window, a token vote.
2. [Limitless](https://predictionmarkets.tools/tools/limitless.md) — Wallet-only on Base, fully collateralised in USDC, and anyone can open a market and take a creator fee — permissionless creation with real collateral behind it.
3. [Myriad](https://predictionmarkets.tools/tools/myriad.md) — An AMM across four chains, settled in stablecoins, wallet-only, 0.5 percent to the taker — the shape of Augur's own pools, on chains with live markets.
4. [Futuur](https://predictionmarkets.tools/tools/futuur.md) — The broadest question list in this catalogue, in USDC or USDT with no identity check, resolved by an editorial team that names its source for every market.
5. [Manifold](https://predictionmarkets.tools/tools/manifold.md) — Anyone opens a question and the creator resolves it, on MIT-licensed code you can read — Augur's open, permissionless half, minus the money.

## The product

Augur is the oldest thing in this category. It held one of Ethereum's first crowdsales in 2015 and
launched on mainnet in 2018 — those two dates are the project's own, from its blog post of 21 July
2026 — and it was built on a position that was unusual then and is still rare: no company and no
committee gets the final word on what happened. A market named a reporter who posted the first
outcome after close; anyone who disagreed could stake REP against it; each round of disagreement
cost more than the last; and if the dispute escalated past every round, the protocol itself split.

That last step is the reason this page exists in 2026 rather than 2021.

**The mechanism ran.** On 8 April 2026, a community member — Micah Zoltu, who had crowdfunded the
attempt the previous August — pushed a dispute over whether the Artemis II mission launched
successfully far enough to trigger a fork. An escalation game ran from April into early June. Then
the protocol did what its whitepaper says it does: it created one child universe per possible
outcome, and every REP holder had 60 days to move their tokens into the one they believed was
right. The window ran from 4 June 2026 at 01:00 UTC to 3 August 2026 at 01:00 UTC.

The numbers are on chain and we checked them against the announcement rather than taking it on
trust. Of the original 11,000,000 REP, 6,547,546.66 migrated — about 59.5 percent. The
`REPv2_Yes_1` token at `0xCf6A0A7826fa124B7705d6f3c675eAD76f1e540D` reports a total supply of
6,545,760.433666705616389974 and the `REPv2_No_1` token at
`0x2F4005456c2F098358213f01DbE34abDAa2989A4` reports 1,786.2274008665274. Both figures match the
announcement to the decimal. The pre-fork REPv2 contract still reports 2,697,575.656086505648896712
left behind in the parent universe, and the 2015-era REPv1 contract still reports its original
11,000,000. The parent universe and both children all hold code and all still respond.

So the protocol is not dead, and a page that called it dead would be wrong. **What does not exist
is an application.** The project's own FAQ, read on 19 September 2026, puts it plainly: not as a
new Augur-native prediction-market application, because one has not yet relaunched, and the site is
not a trading interface. Everything `www.augur.net` serves today is a reboot site — a mission page,
a blog, a whitepaper library, a learn section and a link to `v3.augur.net`, which is ForkWatch, a
read-only archive of the migration. `app.augur.net`, `docs.augur.net` and `blog.augur.net` do not
resolve at all.

The on-chain record says the same thing more quietly. In the 40,000 blocks to 19 September 2026 —
roughly five and a half days — the Augur v2 registry contract at
`0x23916a8F5C3846e3100e5f587FF14F3098722F5d` emitted 1,101 events, and they are almost entirely
bookkeeping: 731 REP balance changes, 365 token transfers, one mint, one reporting-fee
recalculation, one dispute window opened on schedule, one share-token balance change, and one
address claiming the proceeds of a settled position. No market was created. No outcome was
reported. No dispute was contributed to. Of the 731 balance changes, 605 were in the parent
universe — the one the project says is expected to be worthless — and 126 in the surviving one; we
can see that they happened and we cannot tell you what they were.

**This is why the header on this page says the terms changed rather than shut down.** Every other
subject in this section stopped. Augur did not: it is between versions, with the old one intact and
unusable and the new one not deployed. A foundation, Lituus, has funded two teams since March 2025.
One works with ChainSafe on a redesigned oracle, in a repository last touched on 17 September 2026.
The other, Dark Florist, is building a successor protocol and three interfaces in the open — a
repository called `zoltar`, under the Unlicense, whose most recent commit on the day we checked was
that same day. It carries a file of mainnet deployment addresses; we looked up all fifteen, and
fourteen hold no code on Ethereum mainnet. The successor is real, it is public, and it is not
deployed.

Meanwhile the token trades. Kraken lists the migrated token as AUGUR and its AUGUR/USD pair was
online on 19 September 2026, last trade $0.8546, with 1,228 tokens changing hands over the previous
24 hours. On Uniswap the largest pool held about $158,000 and did about $222 in the same day.
CoinGecko renamed its old entry to "Augur [OLD]" and still prints a price for the stranded token —
around $1.05, above the live one, on volume of a few hundred dollars from one Indonesian exchange
and two old Uniswap pairs. A price on a token nobody can use is not a market; it is a listing
nobody has got round to removing.

## What to use instead

Augur was four things in one protocol — a no-permission market factory, a self-custody venue, a
dispute system nobody owned, and a codebase you could read instead of a terms page. They do not
land in the same place.

**If what you had was a wallet and no identity check**, [Polymarket](https://predictionmarkets.tools/tools/polymarket) is the
closest working venue and the closest in design. An account is a wallet, there is no identity check
in the ordinary flow, and outcomes are proposed to the UMA optimistic oracle with a bond behind
them and a two-hour challenge window; a proposal disputed twice escalates to a vote of UMA token
holders. That is the only resolution process left in this catalogue that a third party can enter
with their own money, which was the whole of Augur's argument. It is not the same argument: the
voters hold a different token, there is no universe split behind them, and Polymarket blocks the
United States outright. [Limitless](https://predictionmarkets.tools/tools/limitless) is the same accessibility with a plainer
mechanism — YES/NO shares fully collateralised in USDC on Base, an order book rather than a curve,
1.5 percent to the taker, no identity check.

**If what you had was the habit of opening your own markets**, Limitless is the only pick here
where you can do that with real collateral and take a creator fee on it, which is the closest thing
to Augur's market-creation bond and creator-fee pair. [Manifold](https://predictionmarkets.tools/tools/manifold) is the
unrestricted version and comes with the obvious caveat: anyone opens a question, the creator
resolves it, the code is MIT-licensed and readable, and the currency buys nothing. If you were
creating markets to find out what people thought, that is a small loss. If you were creating them
to earn the fee, it is the whole thing.

**If what you want is the mechanism Augur used**, [Myriad](https://predictionmarkets.tools/tools/myriad) prices with an AMM across
four chains, settles in stablecoins, needs a wallet and nothing else, and takes 0.5 percent — the
cheapest here. Augur v2's own liquidity was pooled the same way. What Myriad does not have is
Augur's back half: the team resolves every market and there is no challenge path.

**If you simply want a venue with a lot of questions that will still be there next year**,
[Futuur](https://predictionmarkets.tools/tools/futuur) runs the broadest subject list in this catalogue, in USDC or USDT alongside
a play-money book on the same questions, with no identity check and an editorial team that names
the public source each market resolves against.

**If you are in the United States**, none of the above is open to you, and Augur effectively never
was either — there is no US-regulated venue in this catalogue that resolves the way Augur does,
because resolution by a token-holder vote is not something a designated contract market does. The
regulated answers are in the [prediction market venues](https://predictionmarkets.tools/categories/prediction-market-venues)
category, and they are a different product with a different bargain: dollars, an identity check,
and a rulebook filed with a regulator instead of a fork.

**And if what you actually wanted was a protocol rather than a product**, nothing here replaces it,
and it is worth saying so rather than pretending. Four of the five picks are hosted businesses with
terms pages. The fifth pays in play money.

## What you give up

**A market you can open without asking anyone.** Augur's `Universe` contract exposes
`createYesNoMarket`, `createCategoricalMarket` and `createScalarMarket` to any address willing to
post the bonds, with the creator naming the fee and the reporter. Limitless has creator markets
with a creator fee and is the nearest working equivalent; Manifold is unlimited and pays in mana.
Everywhere else in this catalogue an operator decides which questions exist, and a question nobody
at the venue wants is a question that does not exist.

**Scalar and categorical markets.** Every venue in the list above is binary YES/NO. Augur v2 had
markets with up to eight named outcomes and scalar markets whose payout is a number in a range
rather than one of two answers. If those are what you were using, you are not changing venue, you
are changing what you can express — and the honest first step is to work out whether the question
you care about can survive being rewritten as yes or no.

**Resolution that nobody can overturn, and the fork behind it.** This is the real loss and 2026 is
the year it became demonstrable rather than theoretical. On Myriad and Futuur a team resolves and
there is no challenge mechanism at all. On Limitless a price feed resolves the automated markets
and the team resolves the rest. Polymarket's UMA path is a genuine dispute process a stranger can
join, and it is still a bonded proposal escalating to a token vote — not a stake-weighted
escalation game ending in the protocol splitting in two and every holder having to choose. Augur
ran that to completion this year. Nothing you can trade today implements it.

**Rules compiled rather than published.** The v2 contracts are MIT-licensed and public, which means
the bond sizes, the fee caps, the dispute rounds and the fork threshold were constants you could
look up and disagree with. What the replacements charge and how they resolve is a claim on a page,
and a page can be edited between your reading it and your position settling.

**What you get back is that there is something to trade.** Every venue above is currently matching
orders, taking deposits and paying out. That sounds like faint praise until you have spent an
afternoon confirming that a protocol works perfectly and has no markets in it.

## Migration notes

**If you did not migrate by 3 August 2026, the tokens are still yours and the door is shut.** The
deadline lives in the fork contract, not in a policy, which means there is no support address that
can make an exception and no governance vote that can reopen it. Your REPv1 or REPv2 still sits in
your wallet and still transfers. The project says it expects it to be worthless. Read every offer
of late migration, recovery or a replacement claim as an attempt to take what is left, because both
the project and its own ForkWatch page say exactly that, unprompted.

**Check the contract address before you touch anything called REP.** There are now four live
tokens with nearly the same name: REPv1 from 2015, REPv2 from the v2 migration, and one child token
per fork outcome. Exchanges do not use the on-chain symbols — Kraken lists the surviving token as
AUGUR — so the address is the only thing that disambiguates them. The one development continues on
is `0xCf6A0A7826fa124B7705d6f3c675eAD76f1e540D`. A wallet that shows you nothing is usually just a
wallet that has not added the new token yet; `0x2F4005456c2F098358213f01DbE34abDAa2989A4` is the
losing side and holds 1,786 tokens in total.

**Where it was custodied decided it for you.** ForkWatch's archived record lists Kraken, OKX and
Bitpanda as supporting the fork and Coinbase, Gate.io and Upbit as not. Kraken's notice, last
updated 17 July 2026, describes a 1:1 migration of custodial balances into the surviving universe
with no client action required, trading paused on 24 July and AUGUR credited by the end of 31 July
— and adds that clients in the EEA and Canada could withdraw the new token but not trade or deposit
it, pending a MiCA whitepaper filing. On 19 September 2026 Coinbase's public API still described
REP-USD as delisted with trading disabled, against the pre-2020 token contract. If you held REP
somewhere and have never seen a new balance, that exchange is the only party who can tell you what
happened to it.

**Anything you built that reads Augur is reading a fork it does not know about.** A dashboard, a
bot or a notebook pointed at the v2 contracts will keep working and keep returning numbers, because
the contracts are live. What changed underneath is which universe those numbers describe and which
token address is the current one. Nothing errors. Make the token address an assertion at startup
rather than a constant somebody set in 2021.

**Do not follow the old documentation.** The v2 repositories are not archived, which reads as
health and is not: the client's last published release is v2.1.13 from 5 March 2021 and the last
commit on the main repository is from 23 July 2023. The GitBook help centre still returns 200 and
still explains how to sign up, add funds, trade, read a portfolio and create markets, and still
presents Augur Turbo on Polygon as a live product. Treat all of it as history.

**Nothing transfers to any of the venues above.** There is no snapshot, no partner venue and no
recovery token. Every pick here is a new account on a different chain funded with different
collateral, and the contracts are not the same contracts. Anyone telling you otherwise is selling
you the replacement.

**Re-check before you act on this page.** This is the one subject in this section where the
direction of travel is towards working rather than away from it. A funded team is writing the
successor protocol in public and has a mainnet address file ready for contracts it has not deployed
yet. Everything above describes 19 September 2026; if you are reading it much later, the first
thing to check is whether those addresses have code.

## FAQ

### Can I trade on Augur today?

No. The project's own FAQ, read on 19 September 2026, answers this directly — not as a new Augur-native prediction-market application, because one has not yet relaunched, and the site is not a trading interface. The v2 contracts are still on Ethereum and still emitting events, but nothing the project publishes puts a market listing in front of you, and we found no third-party interface that does either.

### Is Augur shut down?

No, and this is where most of the older articles are wrong in the other direction. In 2026 the protocol ran its dispute mechanism end to end for the first time, a foundation is funding two teams to rebuild it, and a successor codebase had commits on the day we checked. What does not exist is a place to trade. Shut down and unusable are not the same claim, and Augur is the second one.

### I held REP. What happened to it?

On 3 August 2026 at 01:00 UTC a 60-day migration window closed, and REP had to be moved into one of two child universes before then. About 59.5 percent of the 11,000,000 supply moved, almost all of it into the universe whose token is REPv2_Yes_1 at 0xCf6A0A7826fa124B7705d6f3c675eAD76f1e540D. REP that stayed behind is still in your wallet, still transferable, and the project states plainly that it is expected to be worthless.

### I missed the migration. Is there anything I can do?

No. The deadline was enforced by the fork contract rather than by a policy someone can waive, so it cannot be reopened and nobody can migrate on your behalf. The project and its ForkWatch page both say the same thing and both add the same warning, which is worth repeating — anyone offering late migration, recovery or a replacement claim is running a scam.

### My REP was on an exchange. Did it move?

It depends which one. ForkWatch's archived record lists Kraken, OKX and Bitpanda as having supported the fork and migrated customer balances, and Coinbase, Gate.io and Upbit as not having supported it. Kraken's own notice, last updated 17 July 2026, says it migrated custodial balances 1:1 into the surviving universe with no action required, and lists the result as AUGUR. On 19 September 2026 Coinbase's API still returned a REP-USD product marked delisted with trading disabled, pointing at the pre-2020 token contract.

### Why does the old Augur documentation still describe a working trading app?

Because it was never taken down. The GitBook help centre still returns HTTP 200 and still walks you through signing up, adding funds, trading, reading a portfolio page and creating markets, and still describes Augur Turbo on Polygon as a current product. None of that reflects what exists in September 2026. It is the single most misleading page a reader arriving from an old link will land on.