NFT auction models: english, dutch, and timed reserve explained

05 September 2026

The way an NFT sells depends entirely on how the auction is structured. Three models dominate the market: English auctions, Dutch auctions, and timed reserve auctions. Each one changes who sets the price, how bidders behave, and when the sale settles. Understanding the difference matters whether you are minting a primary drop or flipping on the secondary market.

English auction (ascending price)

This is the standard model most people know. The auction starts at a low price, bidders raise it in increments, and the highest bid wins when time runs out or no one raises further. Price discovery moves upward. Bidders reveal their maximum willingness to pay as the clock ticks.

On Foundation, timed English auctions are the default. A listing runs for 24 hours. If a bid arrives in the last 15 minutes, the clock extends by 15 minutes, which prevents sniping. The final price is whatever the last bidder offered. Sellers benefit from competitive tension; buyers must decide fast whether to top the current bid or walk away.

The model works best for unique, high-demand NFTs. It rewards hype. A seller with a strong community can extract more value than a fixed price would allow, but the auction can also fizzle. If only one person bids, the sale closes at that single bid. No floor, no reserve.

Dutch auction (descending price)

A Dutch auction starts high and drops at set intervals. The first bidder to accept the current price wins. Price discovery moves downward. The seller picks a starting point and a decline rate, and bidders wait for a price they like, but risk someone else taking it first.

This model is common for NFT mint launches. Projects use it to let the market find a fair price without a fixed mint cost. If the drop starts at 1 ETH and ticks down by 0.1 ETH every ten minutes, the first person to hit "buy" at 0.7 ETH gets the NFT, and everyone else pays the same declining rate until supply runs out.

Bidder behavior flips versus an English auction. Here, patience is punished. The early bird pays more; the last mover might get a bargain - or nothing if supply sells out. Sellers use Dutch auctions to guarantee a sale, because the price keeps falling until someone bites. It also avoids gas wars during mints, since transactions happen one at a time.

Timed Reserve Auction

This is a hybrid. The auction runs for a fixed period, with a secret reserve price sitting below the visible starting bid. If no bid meets the reserve, the NFT does not sell. If bids cross the reserve, the highest bidder wins when the timer ends.

Sellers use reserves to protect against lowball offers. Bidders see activity but not the floor. On platforms like OpenSea, sellers can set a reserve when listing a timed auction; the reserve is hidden, so bidders must guess how close they are. This adds uncertainty. Some bidders overpay, others underbid and lose the NFT.

The model suits secondary sales where the seller wants a minimum but still wants competitive bidding. It is less common for primary drops, where Dutch or English models are preferred.

Which model for which situation

Primary drops favor Dutch auctions. They let the market set a price without a fixed mint cost, they avoid gas bidding wars, and sellers can clear a whole collection in minutes.

Secondary sales favor English auctions. Hype drives the price up, timed extensions prevent sniping, and unique pieces get their true market value.

Timed reserve auctions sit in between. They protect the seller from a bad outcome while still allowing price discovery, and they work for high-value singles where the seller has a clear floor in mind.

No model is always best. A seller must match the auction type to the asset's audience and the desired outcome. A buyer must understand the rules before placing a bid, because the difference between a steal and an overpay is often just the auction format.

Back to nft markets