Just wanna scroll the news? Take the pill 💊
All explainers

What is Solana? The complete guide

15 min read · updated 15 May 2026

Solana is a public blockchain built around two technical ideas — fast, deterministic block times and parallel transaction execution — and one product idea: a single global state that anyone can build on without paying premium gas. It launched in March 2020 and, by most measures that matter (active users, daily transactions, fee revenue, developer activity), is the largest non-EVM smart-contract chain in the world.

This guide is comprehensive on purpose. If you want a five-minute primer, skip to the TL;DR. If you want to actually understand what Solana does, why it's different from Ethereum, what its weaknesses are, and what to watch in 2026 and beyond, read the whole thing.

TL;DR

  • What: A Layer-1 blockchain that prioritises throughput, low fees, and a single shared state — no rollups, no shards.
  • Speed: ~400ms block times, finality in ~13 seconds, sub-cent transaction fees.
  • What runs on it: DEXes (Jupiter, Raydium, Orca), liquid staking (Jito, Marinade), NFT marketplaces (Tensor, Magic Eden), payments (USDC, USDT, USDPT), DePIN (Helium, Hivemapper), memecoins (BONK, WIF, POPCAT), and an emerging wave of AI-agent infrastructure.
  • Token: SOL. Used to pay fees, secure the network via staking (~7% APY), and participate in governance via foundation programs.
  • Biggest criticism: Historical outages (mostly 2021–2023). The Firedancer client and the upcoming Alpenglow consensus upgrade are the engineering response.
  • Best place to start: Get a wallet, bridge in some funds, and use Jupiter for your first swap.

Why Solana exists

Solana was founded by Anatoly Yakovenko, a former Qualcomm engineer who'd spent a decade working on distributed systems for cellular networks. His core insight, written up in a 2017 whitepaper, was that consensus — the slowest part of every blockchain — could be partially decoupled from execution if validators agreed on time before they agreed on order.

That sounds abstract. The concrete version: in Ethereum and Bitcoin, every validator has to communicate with every other validator to agree on what time it is and what order transactions arrived. That round-trip is what limits block times to seconds (Ethereum) or minutes (Bitcoin). Yakovenko's idea was to use a cryptographic clock that any validator could verify independently, so the "what time is it" problem stopped being a coordination problem. He called this Proof of History.

Around that idea, the Solana team built a chain optimised for what banks and exchanges actually want from a settlement layer: predictable latency, high throughput, low cost per operation, and no congestion-driven fee spikes. The pitch in 2018 was "Nasdaq on a blockchain." The pitch in 2026 is closer to "the public version of Stripe's backend."

How the technology works

Solana stacks several engineering primitives together. None of them are individually revolutionary; the combination is what produces the throughput numbers.

Proof of History

Every Solana validator runs a verifiable delay function — essentially a SHA-256 hash chain that takes a known amount of wall-clock time to compute. Each transaction gets hashed into this chain as it arrives, producing a cryptographic timestamp. Validators don't have to negotiate "what time is it"; the hash chain provides a globally agreed clock that any node can verify in parallel.

This means consensus rounds skip the time-ordering step. Validators only need to agree on the validity of transactions, not their order.

Tower BFT

Solana's consensus protocol is a Byzantine fault tolerant (BFT) algorithm built on top of PoH. Validators vote on the canonical fork, and votes are themselves PoH-timestamped, so a validator's voting history is publicly auditable. Voting against the longest fork costs the validator a "lockout" — a doubling time penalty before they can vote again — which makes equivocation expensive.

Sealevel — parallel execution

This is the part Ethereum doesn't have. Every Solana transaction explicitly declares which accounts it will read from and write to. The runtime, called Sealevel, looks at every pending transaction in the block and runs all non-conflicting transactions in parallel across CPU cores. A transaction transferring USDC between Alice and Bob doesn't have to wait for an unrelated transaction minting an NFT to Carol.

The architectural consequence: transaction throughput scales with hardware. As validator CPUs get faster and add more cores, the chain gets faster. Ethereum's serial execution doesn't have this property.

Gulf Stream, Turbine, and Pipelining

Three optimisations most users never hear about but which matter for throughput:

  • Gulf Stream forwards transactions to the next expected leader before that leader's slot begins, so when their slot arrives the mempool is already warm.
  • Turbine breaks blocks into shreds and distributes them across the validator set in a tree-shaped fan-out — like BitTorrent for new blocks. This means the leader doesn't have to upload the entire block to every validator.
  • Pipelining processes different stages of block production (fetch, verify, execute, commit) in parallel on different CPUs, like an assembly line.

Single global state

Unlike Ethereum's roadmap (Layer 2s, sharding, rollups), Solana has explicitly stayed monolithic. Every contract on Solana sees the state of every other contract in the same block. A swap on Jupiter can pull liquidity from Raydium and Orca atomically; a liquidation on Kamino can route through Jupiter and rebalance in a single transaction. This composability is the thing that's hardest to replicate on chains that scale by fragmentation.

What runs on Solana in 2026

The chain has matured into several distinct application categories, each with billions in activity.

DeFi — the trading layer

Solana's DeFi ecosystem is dominated by Jupiter, an aggregator that routes swaps across every major Solana DEX. It's the most-used DeFi app on the chain. Below it are the actual liquidity venues:

  • Raydium — the largest AMM, where memecoin liquidity tends to launch.
  • Orca — concentrated-liquidity AMM with the cleanest UX in the ecosystem.
  • Phoenix — a true on-chain limit order book, popular with HFT-style market makers.
  • Drift and Mango — perpetuals platforms with deep liquidity and self-custody.
  • Kamino and marginfi — money markets where you can lend and borrow.

Track the latest DeFi news for protocol launches and TVL movements.

Liquid staking

You can stake your SOL directly with a validator and earn ~7% APY, but your stake is locked. Liquid staking protocols issue a transferable token (JitoSOL, mSOL, bSOL, INF) that represents your staked SOL plus accumulated rewards. You can then use that token in DeFi while still earning staking yield. JitoSOL is the largest, with the additional twist of earning MEV rewards on top of base staking yield.

NFTs and marketplaces

Solana NFTs went through a brutal cycle in 2022–2024 but the ecosystem has restructured. Tensor and Magic Eden are the dominant marketplaces. The real innovation has been compressed NFTs — a state-compression technique that lets projects mint millions of NFTs for fractions of a cent each, enabling use cases like on-chain music distribution, in-game items, and proof-of-attendance tokens.

Stablecoins and payments

USDC and USDT are the workhorses, with USDC on Solana being the largest non-Ethereum USDC deployment. Several payment-focused stablecoins have launched specifically on Solana — including USDPT from Western Union — taking advantage of Solana's sub-second settlement to enable real cross-border flows.

DePIN — physical infrastructure on-chain

Decentralised Physical Infrastructure Networks pay people in tokens for contributing real-world capacity: wireless coverage (Helium), street-level imagery (Hivemapper), GPU compute (Render, io.net), and storage (Shdw). These need cheap, frequent on-chain writes that simply aren't economical on Ethereum. Solana is essentially the only chain where DePIN works at scale.

Memecoins

Solana became the centre of the memecoin cycle in 2024–2025. BONK, WIF, POPCAT, and dozens of smaller tokens drove enormous transaction volume and onboarded millions of new users. Launchpads like Pump.fun lowered the barrier to launching a token to near-zero. Memecoins are controversial — most lose value — but they're also responsible for a meaningful share of Solana's daily fee revenue.

AI agents and machine commerce

The newest category. AI agents that hold their own wallets and transact autonomously — buying compute, paying for data, swapping assets — need a chain with sub-second latency and predictable fees. The Solana Foundation announced a partnership with Google Cloud in 2026 explicitly aimed at this: stablecoin payments built into agent infrastructure. It's early but moving fast. See the Solana news category for the latest in this space.

The economics of SOL

Supply and inflation

SOL has no hard cap. Its supply is managed by a tapering inflation schedule that started at 8% in 2020 and decreases by 15% annually, asymptoting toward a long-term rate of 1.5%. As of 2026, inflation sits around 4.5% per year. Stakers receive almost all of this issuance.

Fees and the burn

Every Solana transaction has a tiny base fee (around 5,000 lamports, where 1 lamport = 0.000000001 SOL). 50% of every base fee is burned, removing SOL from circulation; the other 50% goes to the validator. Priority fees (paid to bid for inclusion in congested blocks) flow entirely to the validator.

At Solana's current transaction volume, burns from base fees roughly counterbalance inflation issuance for a substantial portion of new SOL — meaning the chain is approaching net-neutral supply at sustained activity levels.

Staking economics

Stakers earn ~7% APY from base inflation. Validators take a commission (typically 5–10%), so delegators net 6–6.5%. Running a validator is expensive — bare-metal hardware, dedicated bandwidth, ongoing infra — costing roughly $30k–60k/year. That cost concentrates the validator set among professional operators, which is one of the chain's persistent decentralisation concerns.

What it actually costs to use

  • Average transaction: $0.0005 – $0.005 (a fraction of a cent)
  • Block time: 400 milliseconds
  • Finality: ~13 seconds (32 confirmed blocks)
  • Throughput: ~3,000 TPS sustained in practice; theoretical limit much higher

Solana vs Ethereum (the honest comparison)

The framing wars get tedious, but the technical distinctions matter:

  • Throughput. Solana sustains ~3,000 TPS on mainnet. Ethereum mainnet does ~15 TPS; its Layer 2 rollups together handle ~100–200 TPS.
  • Fees. Solana: fractions of a cent. Ethereum L1: $1–$50 depending on congestion. L2s: $0.01–$0.50.
  • Composability. Solana keeps single-chain composability. Ethereum's roadmap fragments liquidity and state across rollups, requiring bridges or atomic-cross-rollup primitives that don't yet exist in production.
  • Decentralisation. Ethereum has more validators (~1M+ vs Solana's ~1,500). Solana validators are more expensive to run but the chain is fully open-source and permissionless.
  • Stability. Ethereum has never had a full network outage. Solana has had several historically, though none in 2025 or 2026 to date. Firedancer (an independent validator client written from scratch by Jump Trading) is shipping in stages and meaningfully improves client diversity.
  • Use cases. Ethereum is the home of "store of value DeFi" — high-value, infrequent transactions. Solana is the home of "consumer DeFi" — high-frequency, low-value transactions where fee floor matters more than maximum finality assurance.

Both chains will exist for a long time. They optimise for different things.

Criticisms worth knowing

Network outages

Solana experienced multiple multi-hour outages between 2021 and early 2023, mostly triggered by transaction floods from automated NFT mint bots or runaway oracle workloads. The chain has had no full outage in 2024, 2025, or 2026 to date. The fixes have been gradual: priority fee markets to throttle abusive workloads, QUIC for validator-to-validator transport, ongoing client diversity work via Firedancer, and the upcoming Alpenglow consensus upgrade.

Validator hardware

Running a Solana validator requires enterprise-grade hardware — at minimum 256 GB RAM, NVMe storage, 1 Gbps bandwidth, and 12+ CPU cores. That's roughly $30k/year in infrastructure costs, before staff time. This concentrates the validator set. The counter-argument from the foundation is that what matters is the number of independent operators, not the consumer-hardware-feasibility of the role.

The FTX legacy

Sam Bankman-Fried was the most prominent Solana booster pre-2022. Alameda Research held a substantial SOL position and FTX was a major venue for SOL trading. The chain survived FTX's collapse intact — no protocol-level damage, no broken validator set — but the association lingered for years and still surfaces in critical coverage. Solana's response has been operational: ship more, sponsor more independent research, build a broader corporate user base (Visa, Western Union, Google Cloud).

Memecoin reputation

Solana's dominance in the 2024–2025 memecoin cycle was a double-edged sword. It drove massive activity but also painted the chain as a "casino" in mainstream coverage. The institutional integrations of 2025–2026 (Western Union, BlackRock-affiliated funds, State Street's onchain offerings) have shifted that perception, but slowly.

Upgrades to watch

Firedancer

An independent validator client written from scratch by Jump Trading, in C, designed to be radically more efficient than the Rust reference client. A subset of Firedancer has been running in production since 2024 under the name "Frankendancer." The full client is rolling out in stages. Once a meaningful portion of stake runs Firedancer, Solana will have true client diversity for the first time — eliminating an entire class of outage risk.

Alpenglow

Solana's biggest consensus upgrade in its history, expected to land in 2026. Replaces the current voting protocol with one designed for sub-second finality at the chain's current throughput. Co-founder Anatoly Yakovenko has called it the "single biggest change since launch." Read the latest Alpenglow updates.

Token Extensions and confidential transfers

An evolution of Solana's token standard that adds programmable features — transfer hooks, interest accrual, confidential balances — to SPL tokens. Crucial for institutional and regulated stablecoins (Western Union's USDPT and several CAD/EUR-pegged stablecoins use Token Extensions).

How to actually use Solana

If this is your first time, the practical sequence:

  1. Install a wallet. Phantom is the most popular; Backpack and Solflare are strong alternatives. See our wallet guide.
  2. Get some SOL. Either buy directly on Coinbase / Binance / OKX / Kraken and withdraw, or bridge in from Ethereum.
  3. Make your first swap. Use Jupiter to swap a small amount of SOL into USDC or another token. Fees will be cents.
  4. If you're holding longer-term: stake your SOL or convert it to JitoSOL for liquid staking.
  5. Stay informed. Follow the Solana news feed, or read the weekly report for a curated summary every Monday.

Common questions

Is Solana decentralised?

Less than Ethereum by validator count (~1,500 vs ~1M), more than Bitcoin by geographic distribution. The Nakamoto Coefficient — the minimum number of entities needed to halt the chain — sits around 22 for Solana as of 2026. That's higher than most chains and meaningfully decentralised by any practical definition.

Is SOL inflationary?

Technically yes — about 4.5% per year as of 2026, tapering to 1.5% long-term. In practice, base-fee burns at current transaction volumes offset a substantial portion of new issuance.

What's the catch?

Solana trades client diversity (fewer independent validator clients than Bitcoin/Ethereum) for performance. Firedancer fixes the client diversity issue. It also requires more expensive hardware to validate, which concentrates the validator set. These are real engineering trade-offs, not marketing problems.

Will Solana survive long-term?

It has the largest non-EVM developer base, the largest non-EVM stablecoin float, growing institutional adoption (Western Union, State Street, Google Cloud partnerships), and a consensus upgrade pipeline that's been hitting deadlines. The honest answer: barring a critical protocol failure, yes, it survives the next cycle. Whether it takes share from Ethereum or coexists is the open question.

Read more

Learning Solana? We write it up daily.

Solana news in plain English — no jargon walls, no shilling. Free, unsubscribe anytime.

More explainers