What Is Kaspa (KAS) and How Does It Work?

What Is Kaspa (KAS)?

Kaspa is a permissionless Layer 1 network that uses Proof of Work. Permissionless simply means that anyone can use the network or take part in it without asking for permission first.

KAS is the native cryptocurrency of the Kaspa network. Miners receive KAS as a reward for producing blocks and can also receive the transaction fees that users pay.

Kaspa does not use a traditional, linear blockchain where blocks are placed one after another in a single line. Instead, the network uses a BlockDAG. This lets multiple blocks be created at about the same time while still remaining part of the network. On a traditional blockchain, usually only one competing block becomes part of the main chain.

To place these blocks in the right order, Kaspa uses GHOSTDAG. This protocol determines how parallel-produced blocks are ordered and how the network reaches consensus about them.

The Kaspa mainnet went live on November 7, 2021 through a fair launch. There was no premine, presale, or amount of KAS reserved ahead of time for insiders. KAS entered circulation through mining from the start.

Since the Toccata hard fork on June 30, 2026, Kaspa also supports programmable UTXOs through covenants. A UTXO is, in short, a separate piece of crypto that can still be spent. A covenant adds rules that determine how that UTXO may be spent in a future transaction.

This makes it possible to build more complex programmable logic on Kaspa. It just works differently than on an account-based smart contract platform like Ethereum. Kaspa remains UTXO-based: the rules are tied to UTXOs and transactions, instead of accounts and contract storage like on the Ethereum Virtual Machine.

Key Takeaways

• Kaspa is a Layer 1 network with proof of work and the native coin KAS.
• It uses a BlockDAG instead of one linear blockchain.
• GHOSTDAG orders blocks and transactions created in parallel.
• The network runs at 10 blocks per second.
• Proof of work, probabilistic finality, and higher node requirements are important trade-offs.

How Does Kaspa Work?

Kaspa works with miners. These are participants who use computing power to produce new blocks. That process is called Proof of Work. Miners then spread these blocks through a peer-to-peer network, where computers communicate directly with each other without a central party.

On a traditional blockchain, a new block usually points to one previous block. On Kaspa, a block can point to multiple earlier blocks. This lets blocks created at about the same time by different miners exist side by side, instead of only one of them becoming part of the main chain.

GHOSTDAG orders these parallel-produced blocks. This creates a consistent order in which transactions can be processed. That matters, for example, when two transactions try to spend the same UTXO. In that case, the order determines which transaction gets accepted first, and the conflicting transaction can no longer be executed afterward.

Since the Crescendo upgrade, Kaspa has produced an average of 10 blocks per second. That lowered the target block time from about 1 second to 100 milliseconds.

A short block time does not mean every transaction is fully certain after just one block. Certainty in a Proof-of-Work network is probabilistic: as more blocks are added and a transaction gets deeper into the ordered history, certainty increases.

Miners receive KAS through block rewards and transaction fees. Within GHOSTDAG, blocks are classified in part as blue and red blocks. Red blocks are still part of the BlockDAG, but their reward works differently: when such a block is included by a later block, the related reward can go to the miner of that merging block.

Kaspa (KAS) Overview

What Is Kaspa's BlockDAG?

A BlockDAG is a directed acyclic graph of blocks. That sounds technical, but the main difference from a traditional blockchain is simple: the blocks do not form one long chain.

On a regular blockchain, a new block usually follows one previous block. When two miners create a valid block at nearly the same time, usually only one of those blocks eventually becomes part of the main chain.

Kaspa works differently. Multiple valid blocks can exist in parallel side by side, and a new block can point to multiple earlier blocks. That means Kaspa does not automatically have to leave parallel-produced blocks out of history.

You can compare it to a road network with multiple lanes instead of one narrow track. Different blocks can appear at about the same time, after which the protocol decides how to order those blocks logically relative to each other.

The connections between blocks first only show a partial order. For some blocks, it is clear which block came earlier, but not every pair of parallel-produced blocks has a fixed position relative to each other right away.

GHOSTDAG then turns this structure into a fixed order in which blocks and transactions can be processed. That is needed, for example, when two transactions try to spend the same UTXO. The transaction that is processed first according to the GHOSTDAG order gets priority.

Because Kaspa can keep and order parallel-produced blocks, the network can work with a higher block frequency than a design that follows only one block path. That does not mean every transaction is automatically valid. Transactions still have to follow the network rules, and conflicts still have to be resolved according to the established order.

How Does GHOSTDAG Work on Kaspa?

GHOSTDAG is Kaspa's consensus mechanism. It makes sure that participants in the network eventually follow the same order of blocks and transactions.

GHOSTDAG is based on PHANTOM. The exact PHANTOM solution is very hard to calculate efficiently. So GHOSTDAG uses a practical method that step by step determines how blocks are classified and ordered.

In that process, blocks are classified as blue or red, among other things. Blue blocks fall within the amount of parallel-produced blocks that the protocol considers normal. Blocks outside that range are classified as red.

An important role in this is played by the value k. It determines how many simultaneously produced blocks may fall within the blue group. That value is tied in part to the block frequency and the time blocks need to spread through the network.

GHOSTDAG then chooses a so-called selected parent step by step. This is an earlier block whose past contains the largest amount of accepted blue work according to the protocol. The order is then built further from that selected parent.

Kaspa uses blue work for this, among other things. This is a measure of the amount of Proof of Work connected to the accepted blue blocks. It helps GHOSTDAG determine which branches and blocks carry more weight in the final order.

The certainty of that order increases as more blocks are added. The longer a transaction is part of the ordered history, the harder it becomes to change that order.

GHOSTDAG is currently the consensus protocol of the Kaspa mainnet. DAGKnight has been developed as a possible next consensus upgrade. One important difference is that DAGKnight is not supposed to depend on one preset k value, but should be able to adapt better to the network's actual conditions. DAGKnight is not yet active as a consensus mechanism on the Kaspa mainnet.

Who Developed Kaspa?

Yonatan Sompolinsky is the founder of Kaspa. Before Kaspa launched, he was already working on research into ways a Proof-of-Work network can process and order multiple parallel-produced blocks.

Together with Shai Wyborski and Aviv Zohar, he published the research on PHANTOM and GHOSTDAG in 2021. This research forms an important basis for the way Kaspa orders blocks within a BlockDAG.

Kaspa was originally developed by DAGLabs. Later, the project was handed over to the community and continued as an open-source project.

Michael Sutton is one of the main developers and researchers within Kaspa. Kaspa's original full node software was written in Go. That software was later rebuilt in Rust to allow better performance.

Michael Sutton and Ori Newman played an important role in this Rust version. In addition, several other open-source developers help with Kaspa's ongoing development.

What Are the Benefits of Kaspa?

Kaspa has a number of technical features that set it apart from a traditional Proof-of-Work blockchain:

• Parallel blocks are processed. Kaspa can include and order blocks that are mined at about the same time. On a traditional blockchain, usually only one block path is followed.

• High block frequency. The Kaspa mainnet runs at about 10 blocks per second. The target time between two blocks is therefore about 100 milliseconds.

• Fixed transaction order. GHOSTDAG determines a fixed order for blocks and transactions. That lets nodes follow the same history and process conflicting transactions in the same way.

• Fair launch. Kaspa launched without a premine, presale, or amount of KAS reserved ahead of time for insiders. New KAS entered circulation through mining from the start.

• Programmable UTXOs. Since the Toccata upgrade, Kaspa supports covenants. These let rules be attached to the way a UTXO may later be spent. Toccata also added new ways to use information from transactions and verify certain zero-knowledge proofs directly on Layer 1.

A high block frequency does not automatically mean every transaction is cheap or instantly final. Actual performance also depends on things like network congestion and how much confirmation a user wants.

What Are the Downsides of Kaspa?

Kaspa also has a number of downsides and technical trade-offs:

• Dependence on miners. Kaspa uses Proof of Work. The network's security therefore depends on miners providing computing power.

• Transactions are not instantly absolutely final. As with other Proof-of-Work networks, a transaction becomes more certain as more blocks are added. The longer a transaction stays in the ordered history, the harder it becomes to change that order.

• The value k has to be set correctly. GHOSTDAG uses a value called k. This determines how many parallel-produced blocks may fall within a certain group. That value has to fit things like how fast blocks are created and spread.

• Nodes need relatively powerful hardware. Because of the high block frequency and the amount of data the network processes, nodes need enough computing power, memory, storage, and internet bandwidth.

• Network upgrades require action from node operators. During major protocol upgrades, operators need to update their node software on time. A node that keeps using old rules may no longer communicate correctly with the updated network after such an upgrade.

According to the current guidelines for the Kaspa node, a minimum of 8 CPU cores, 16 GB RAM, 640 GB SSD storage, and about 80 Mbit/s of network bandwidth is recommended. These are practical minimum recommendations for the current software and may change with future upgrades.

Conclusion

Kaspa is a Proof-of-Work Layer 1 network with KAS as its native cryptocurrency. The main difference from a traditional blockchain is the BlockDAG structure. This lets multiple valid blocks appear at about the same time while still being part of the same history.

GHOSTDAG determines how these blocks and the related transactions are ordered. That lets Kaspa currently process about 10 blocks per second without being limited to one linear block path.

Since the Toccata upgrade, Kaspa can also run more programmable logic directly on Layer 1 through covenants and other new features.

There are trade-offs too. Kaspa still depends on miners for security, transaction certainty increases gradually, and running a node requires relatively powerful hardware. So Kaspa is mainly a different technical approach to making a Proof-of-Work network work faster without automatically throwing away parallel-produced blocks.