Glossary · Blockchain development

Testnet

Published 24 May 2025 · Updated 21 Oct 2026 · 13 min read

A testnet is a separate blockchain for testing software and upgrades before deployment to mainnet, the live blockchain. It shares software and most rules, but has its own history, chain ID and coins that are not supposed to have value. Developers rehearse changes here, although conditions differ from those of a live chain.

  • A testnet shares most rules, but has its own chain ID and coins intended to have no value.
  • Five major Ethereum upgrades since 2022 reached mainnet 43 to 99 days after their first long-lived testnet.
  • Testnets get replaced, with Sepolia for Ethereum apps and Hoodi for validators in October 2026.
How a testnet differs from mainnetTestnets and other testing environmentsHow upgrades move through testnetsPublic testnets in 2026Bitcoin's test networks and the 20-minute ruleTest coins, faucets and incentivized testnetsWhat a testnet cannot showFrom rehearsal to production deploymentTestnet FAQ

How a testnet differs from mainnet

Shared client software does not mean shared transactions or balances. A transfer recorded on one chain does not update another. These test networks also differ in who validates blocks, how much computation blocks allow and how busy they are. Those differences matter when interpreting a successful testnet deployment or comparing transaction costs.

Separate chain identity

Each testnet starts from a genesis block separate from the main chain's. Ethereum uses a chain ID for replay protection. Bitcoin uses network magic and prefixes such as m, n, 2 and tb1.

Coins without intended value

These coins exist only on their own chain and are supposed to have no value. Distribution can involve faucets or mining, depending on the network.

Different validator membership

In October 2026, Sepolia has a permissioned validator set. Hoodi allows anyone to run a validator and test proof of stake operations.

Different settings and activity

Sepolia versus mainnet measurements from 21 October 2026 show 16 wei versus 0.69 gwei base fees and 200 million versus 60 million gas limits. Hoodi misses about 9.3% of slots, versus mainnet's 0.29%.

Testnets and other testing environments

Testing does not always require a public chain. Hardhat Network runs locally with chain ID 31337. Anvil creates ten development accounts with 10,000 ETH each and can fork a chain at a chosen block. Other environments preserve network history or involve additional operators. The right distinction is what the environment can reproduce, not simply whether it calls itself a testnet.

EnvironmentWho runs itState and coinsGood for
Local development chainWho runs itA developer on one machineState and coinsLocal state and separate development coinsGood forIsolated contract tests and controllable block creation
Mainnet forkWho runs itTeams running local or hosted environmentsState and coinsCopied live state, including contract balances and deployed protocolsGood forInteraction tests with existing contracts, balances and protocols
DevnetWho runs itClient teams preparing a network upgradeState and coinsTemporary upgrade state and separate test coinsGood forEarly checks before long-lived public rehearsals
Shadow forkWho runs itA small set of node operatorsState and coinsSource network state, history and balancesGood forTransaction replays against proposed changes under more realistic conditions
Public testnetWho runs itClient teams or open validator communitiesState and coinsIndependent history and coins without intended valueGood forShared application, staking and protocol rehearsals
Resetting testnetWho runs itEphemery's open community of validator operatorsState and coinsExecution state, consensus state and balances with a 28-day resetGood forRepeatable validator practice without retaining a growing history

An incentivized testnet rewards participants with tokens for work during a test phase. Most teams move code through several of these steps before mainnet deployment. A locally reproducible failure and a successful shared-network rehearsal provide different evidence about the same change. Fork-based testing also depends on source data, with Hardhat requiring a node that provides archival data.

How upgrades move through testnets

Client teams start with devnets, then fork long-lived public testnets one after another. A planned hard fork activates the network upgrade's rules for that rehearsal. Successful results come before a mainnet activation date. From 2022 through October 2026, the five major upgrades span 43 to 99 days between their first long-lived public fork and activation. Early Glamsterdam trials exposed consensus, finality and implementation problems in August and September 2026, according to news reports.

Figure 1 · Six Ethereum upgrades, 2022 to 2026Ethereum upgrade rehearsal pathsChoose an upgrade to compare dated stages, time between forks and the incidents that changed its path.
Upgrade
Fork iconMainnet shadow forksShadow fork11 Apr to 9 Sep 202213 forks of mainnet in 151 daysThe first one found minor issues in Nethermind and Besu.
Flask iconRopstenTestnet8 June 2022First long-lived testnetMerged on a beacon chain launched on 30 May.
Bug iconSepoliaIssue found6 July 202228 days laterDevelopers noted a minor block producer problem.
Bug iconGoerliIssue found11 August 202236 days laterTwo terminal blocks and nodes that had not updated delayed finality.
Globe iconMainnetMainnet15 September 202235 days laterParis at block 15,537,394 ended proof of work.
Long-lived testnets3Ropsten, Sepolia and Goerli
First testnet to mainnet99 days8 June to 15 September 2022
Rehearsals on copies of mainnet13Shadow forks from April to September 2022

The Merge's 2022 path included 13 mainnet shadow forks from 11 April to 9 September and 99 days from Ropsten to activation.

Fork iconMainnet shadow forkShadow forkJanuary 2023Before the public testnetsA few minor technical issues with Geth.
Flask iconZhejiangShort-term1 February 2023Public withdrawal testnetOpened to test withdrawals before Sepolia.
Flask iconSepoliaTestnet28 February 202327 days laterA smooth upgrade, per the Ethereum Foundation.
Flask iconGoerliTestnet14 March 202314 days laterSmooth. The last testnet before mainnet.
Globe iconMainnetMainnet12 April 202329 days laterStaking withdrawals went live at epoch 194,048.
Long-lived testnets2Sepolia and Goerli
First testnet to mainnet43 days28 February to 12 April 2023
Short-term testnetZhejiangWithdrawal tests from 1 February 2023

Shapella's 2023 withdrawal rehearsals included Zhejiang from 1 February. Its first long-lived fork preceded mainnet activation on 12 April 2023 by 43 days.

Bug iconGoerliIssue found17 January 2024First testnetA Prysm bug stopped finality for nearly four hours.
Flask iconSepoliaTestnet30 January 202413 days laterNo issues reported.
Flask iconHoleskyTestnet7 February 20248 days laterNo issues reported.
Globe iconMainnetMainnet13 March 202435 days laterBlobs went live. Goerli was retired a month later.
Long-lived testnets3Goerli, Sepolia and Holesky
First testnet to mainnet56 days17 January to 13 March 2024
Issues found on testnets1Prysm on Goerli, fixed in under 4 hours

Dencun's Goerli testnet fork on 17 January 2024 exposed a Prysm bug, interrupting finality for nearly four hours. A fix restored finality.

Flask iconEphemeryShort-termBefore 14 Feb 2025Resets every 28 daysAlready running Pectra when the testnet dates were announced.
Warning iconHoleskyIncident24 February 2025First long-lived testnetThree clients had the wrong deposit contract address. No finality for about two weeks.
Warning iconSepoliaIncident5 March 20259 days laterAn extra log from the permissioned deposit contract caused many empty blocks for about 6.5 hours.
Rocket iconHoodiNew testnet26 March 202521 days laterLaunched on 17 March to replace Holesky for staking tests.
Globe iconMainnetMainnet7 May 202542 days laterActivated at epoch 364,032.
Long-lived testnets3Holesky, Sepolia and the new Hoodi
First testnet to mainnet72 days24 February to 7 May 2025
Incidents on testnets2Holesky and Sepolia, both from testnet configuration

Pectra reached mainnet on 7 May 2025, 72 days after Holesky's fork. Configuration failures and a new chain extended its rehearsal path.

Bug iconHoleskyIssue found1 October 2025First testnetHigh reorg rates on Lighthouse, with a fix planned. The last upgrade on Holesky.
Flask iconSepoliaTestnet14 October 202513 days laterA few operators failed to update nodes. No major bugs.
Flask iconHoodiTestnet28 October 202514 days laterNo issues reported.
Globe iconMainnetMainnet3 December 202536 days laterBlob parameter forks followed on 9 December and 7 January.
Long-lived testnets3Holesky, Sepolia and Hoodi
First testnet to mainnet63 days1 October to 3 December 2025
Issues found on testnets1Lighthouse reorgs on Holesky

Fusaka's 1 October 2025 Holesky fork exposed Lighthouse reorganization problems. It was Holesky's final upgrade before retirement. A fix was planned for the next release.

Terminal iconDevnets 8, 9 and 11DevnetsAug to Sep 2026Client test networksNews reports describe a consensus bug, non-finality and an EIP-8037 issue.
Flask iconPlatåbergetShort-term20 August 2026Short-term community testnetBootstrapped from public deposits, planned to run a few months.
Flask iconSepoliaTestnet6 October 202647 days laterFirst planned for 3 August. It went live 64 days later.
Clock iconHoodiScheduled26 October 2026Scheduled, 20 days laterScheduled. The gas limit rises to 200 million.
Clock iconMainnetNot setNot decidedNo date yetNo mainnet date as of 16 October 2026.
Long-lived testnets2Sepolia done, Hoodi scheduled
First testnet to mainnetNot setMainnet date not decided
Delay before Sepolia64 daysPlanned for 3 August, live on 6 October

Glamsterdam reached Sepolia on 6 October 2026. Hoodi's fork is scheduled for 26 October 2026. The mainnet date remained undecided on 16 October 2026.

Dates and issues follow Ethereum Foundation posts, news reports and developer call notes. Hoodi's 26 October 2026 activation remains scheduled.

Pectra's Holesky rehearsal failed on 24 February 2025 because Geth, Nethermind and Besu had incorrect deposit contract addresses. A supermajority attested to an invalid block. Finality did not return until 10 March 2025. The inactivity leak forced exits into a year-long queue, making full validator lifecycle tests impractical. Hoodi launched on 17 March 2025 after these testing failures. The replacement gave staking teams a usable rehearsal chain rather than a queue they could not reasonably test through.

Sepolia's Pectra fork on 5 March 2025 hit a different configuration mismatch. Its permissioned deposit contract also emitted an ERC-20 log that clients did not expect. Blocks containing deposits were rejected, leaving many blocks empty for about 6.5 hours. Finality continued. The Ethereum Foundation said this issue could not occur on mainnet. Both incidents show why a rehearsal must check network-specific settings, rather than assuming software prepared for production will work unchanged everywhere.

Public testnets in 2026

Purpose matters more than a familiar testnet name. Ethereum separates application testing from staking practice, while Solana separates app testing from core-release stress tests. The directory records test network roles alongside identifiers and access requirements. Its live readings are dated 21 October 2026, so measured behavior is distinguishable from each chain's stated purpose.

Figure 2 · Test networks by ecosystemChoose a public testing environmentSelect an ecosystem and hide retired chains to compare purpose, access rules, identifiers and dated measurements.
Ecosystem
Show
SepoliaActiveDefault testnet for apps and toolingChain ID11155111ValidatorsPermissioned, run by client and testing teamsLaunchedOctober 2021, merged in July 2022Measured 21 Oct12.14 s per block, 16 wei base fee, 200M gas limitFutureThe EF end date of 30 September 2026 passed. A successor is discussed for 2027
HoodiActiveValidators, staking and upgrade testsChain ID560048ValidatorsOpen set, anyone can run a validatorLaunched17 March 2025Measured 21 Oct13.23 s per block, 9.3% missed slotsEnd of lifeSeptember 2028 (EF) or December 2028 (client README)
EphemeryResetsShort tests on fresh stateChain IDChanges with every resetValidatorsOpenResetsEvery 28 days, iteration 164 in October 2026Node sizeUnder 5 GB on average
MainnetMainnetFor comparisonChain ID1Measured 21 Oct12.04 s per block, 0.69 gwei base fee, 60M gas limitMissed slots0.29% across the last 7,200 blocks
HoleskyRetiredFormer staking testnetChain ID17000Launched28 September 2023EndedDeprecated September 2025, nodes shut down after FusakaWhyThe Pectra incident left a year-long exit queue
GoerliRetiredFormer default testnetChain ID5Launched2019EndedSunset on 13 April 2024WhyTest ETH ran short and traded at up to $1.60 in February 2023
RopstenRetiredFirst long-standing testnet to mergeChain ID3Merged8 June 2022EndedShut down 15 to 31 December 2022
RinkebyRetiredGeth proof-of-authority testnetChain ID4EndedMid-2023WhyIt did not support the Merge
KovanRetiredProof-of-authority execution testnetChain ID42EndedBefore June 2022

In October 2026, Sepolia serves applications and Hoodi supports staking. Ephemery resets every 28 days for repeatable validator practice.

testnet4ActivePublic test network meant to replace testnet3SinceBitcoin Core 28.0, October 2024Height157,211 on 21 October 2026Minimum difficulty90.5% of the last 2,016 blocksDefault port48333
signetActivePredictable test network with signed blocksSinceBitcoin Core 0.21.0, January 2021Height327,352 on 21 October 2026Average block592 s since genesisBlocks needA signature from the signet challenge
testnet3DeprecatedThird public test networkSinceBitcoin Core 0.7, 2012Height5,195,969 on 21 October 2026Average block95 s since its 2011 genesisStatusDeprecated in Bitcoin Core 28.0, still in 31.1
regtestLocalPrivate chain on one machineBlocksMined on commandCoinsBlock rewards spendable after 100 blocksUsed forUnit and integration tests

Bitcoin Core 31.1 still supports testnet3 in October 2026. Its planned phaseout remains incomplete despite testnet4's arrival in October 2024.

Solana DevnetActiveFor application developersRPCapi.devnet.solana.comFaucet2 requests every 8 hoursMeasured 21 Oct238 ms per slot, version 4.4.0-beta.1ResetsMay be subject to ledger resets
Solana TestnetActiveStress tests of new releases by core contributorsRPCapi.testnet.solana.comSoftwareTypically newer than Devnet and MainnetMeasured 21 Oct218 ms per slot, version 4.4.0-beta.1
Canton DevNetResetsIntegration testing, upgraded firstAccessAllowlisted IP, VPN and a Super Validator sponsorResetsAbout every 3 monthsCoinsTest Canton Coin from a faucet
Canton TestNetResetsStaging, upgraded after DevNetAccessTokenomics Committee approval and a Super Validator sponsorResetsEvery 3 to 6 monthsCoinsTest Canton Coin from a faucet
Canton MainNetMainnetProduction, for comparisonResetsNeverCoinsCanton Coin has real valueSize62 Super Validator nodes, 835 validators in total

Solana Devnet hosts apps, and Solana Testnet stress-tests releases. Canton DevNet upgrades before TestNet, which requires Tokenomics Committee approval and Super Validator sponsorship in October 2026.

Arbitrum SepoliaActiveArbitrum One test networkChain ID421614Parent chainSepolia
OP SepoliaActiveOP Mainnet test networkChain ID11155420Parent chainSepolia
Base SepoliaActiveBase test networkChain ID84532FaucetCoinbase CDP API, ETH and USDC
Polygon AmoyActivePolygon PoS test networkChain ID80002Parent chainSepoliaReplacedMumbai, retired on 13 April 2024 with Goerli
Avalanche FujiActiveOfficial Avalanche testnetChain ID43113 (C-Chain)FaucetNeeds mainnet AVAX or a faucet coupon

In October 2026, Arbitrum Sepolia, OP Sepolia and Base Sepolia settle to Sepolia. Amoy replaced Mumbai, which ended on 13 April 2024.

Network documentation and live readings collected by the page team on 21 October 2026 inform this directory.

Growing history and coin shortages can force retirement. Goerli retired on 13 April 2024, and Holesky was deprecated in September 2025. Advice to use either is stale. The retired Ropsten network shut down in December 2022. In October 2026, Sepolia's planned end of life on 30 September 2026 has passed without a new official date. A 2027 replacement is proposed but unnamed. Layer 2 test networks follow their parent testnet.

Bitcoin's test networks and the 20-minute rule

Bitcoin's testnet3 has run since 2012. Its 20-minute rule drops difficulty to the minimum when a block's timestamp exceeds its predecessor's by more than 20 minutes. This permits CPU mining. A minimum-difficulty block ending a difficulty period could also reset the next period's base. That interaction caused a block storm of 12,682 blocks in one day in 2015, versus about 144 normally. In April 2024, Jameson Lopp said he generated over 165,000 blocks in a week for $1 worth of electricity, disrupting testing.

Bitcoin Core added testnet4 in October 2024. It takes the difficulty base from a period's first block and adds a timewarp rule. The 20-minute exception remains, except at a period's first block, and miners still exploit it in October 2026. A Bitcoin Core maintainer described testnet4 as broken in May 2026. Signet adds signatures to proof of work, making block progress more predictable under centralized control. For private testing, regtest lets an operator create blocks whenever needed. Anyone can also run a personal signet.

Figure 3 · Bitcoin test networks, liveBitcoin difficulty and block timestampsChoose a network and move the minutes slider to compare expected work with ten recent blocks and their timestamp gaps.
Network
21 minutes
Expected work for the next block4.3 × 10⁹ hashes
The next block needs the full network difficulty.Drag past 20 minutes. On testnet3 a block stamped more than 20 minutes after its parent drops to difficulty 1.The next block needs the full network difficulty.Drag past 20 minutes. On testnet4 a block stamped more than 20 minutes after its parent drops to difficulty 1.Signet has no 20-minute rule. Each block also needs a valid signature.Only the signers in the signet challenge can produce blocks, so outside miners cannot flood it.Mainnet has no 20-minute rule. Difficulty changes only at a retarget.Every 2,016 blocks the difficulty is adjusted toward one block every 10 minutes.More than 20 minutes have passed, so a block stamped now must use the minimum difficulty.BIP-94 keeps this testnet rule: a block stamped more than 20 minutes after its parent must use difficulty 1, except the first block of a difficulty period.
Latest 10 testnet3 blocks on 21 October 2026, time since the previous block
  • #5,195,960+20 min 1 sMinimum difficulty
  • #5,195,961−1 h 40 min, earlier than parentDifficulty 4,194,304
  • #5,195,962+20 min 1 sMinimum difficulty
  • #5,195,963+20 min 1 sMinimum difficulty
  • #5,195,964+20 min 1 sMinimum difficulty
  • #5,195,965+20 min 1 sMinimum difficulty
  • #5,195,966+20 min 1 sMinimum difficulty
  • #5,195,967+20 min 1 sMinimum difficulty
  • #5,195,968+20 min 1 sMinimum difficulty
  • #5,195,969+1 sDifficulty 4,194,304
Minimum-difficulty blocks36.4%734 of the last 2,016 blocks
Newest block timestamp114 min aheadAgainst real time on 21 October 2026
Average block since genesis95 sTarget 600 s

Readings for 21 October 2026 show minimum difficulty in 36.4% of 2,016 recent blocks. One timestamp precedes its parent's by 100 minutes.

Latest 10 testnet4 blocks on 21 October 2026, time since the previous block
  • #157,202+20 min 1 sMinimum difficulty
  • #157,203+20 min 1 sMinimum difficulty
  • #157,204+20 min 1 sMinimum difficulty
  • #157,205+20 min 1 sMinimum difficulty
  • #157,206+20 min 1 sMinimum difficulty
  • #157,207+20 min 1 sMinimum difficulty
  • #157,208+20 min 1 sMinimum difficulty
  • #157,209+20 min 1 sMinimum difficulty
  • #157,210+20 min 1 sMinimum difficulty
  • #157,211+20 min 1 sMinimum difficulty
Minimum-difficulty blocks90.5%1,825 of the last 2,016 blocks
Newest block timestamp106 min aheadAgainst real time on 21 October 2026
Average block since genesis495 sTarget 600 s

Readings for 21 October 2026 show 90.5% minimum difficulty among the last 2,016 blocks. The latest ten timestamps each advance exactly 20 minutes and 1 second.

Latest 10 signet blocks on 21 October 2026, time since the previous block
  • #327,343+7 min 6 sDifficulty 0.047
  • #327,344+4 min 44 sDifficulty 0.047
  • #327,345+30 minDifficulty 0.047
  • #327,346+7 min 18 sDifficulty 0.047
  • #327,347+11 min 14 sDifficulty 0.047
  • #327,348+4 min 10 sDifficulty 0.047
  • #327,349+1 min 23 sDifficulty 0.047
  • #327,350+23 sDifficulty 0.047
  • #327,351+1 min 40 sDifficulty 0.047
  • #327,352+35 sDifficulty 0.047
Minimum-difficulty blocksNoneSignet has no 20-minute rule
Newest block timestamp2 min behindAgainst real time on 21 October 2026
Average block since genesis592 sTarget 600 s

The 21 October 2026 calculation gives signet a 592-second average block interval since genesis, close to the 600-second target.

Latest 10 mainnet blocks on 21 October 2026, time since the previous block
  • #972,482+2 min 1 sDifficulty 140,577,680,051,920
  • #972,483+6 min 2 sDifficulty 140,577,680,051,920
  • #972,484+11 min 7 sDifficulty 140,577,680,051,920
  • #972,485+9 min 56 sDifficulty 140,577,680,051,920
  • #972,486+3 min 5 sDifficulty 140,577,680,051,920
  • #972,487+2 min 42 sDifficulty 140,577,680,051,920
  • #972,488+11 min 36 sDifficulty 140,577,680,051,920
  • #972,489+1 h 2 minDifficulty 140,577,680,051,920
  • #972,490+3 min 48 sDifficulty 140,577,680,051,920
  • #972,491+2 min 47 sDifficulty 140,577,680,051,920
Minimum-difficulty blocksNoneMainnet has no 20-minute rule
Newest block timestamp1 min behindAgainst real time on 21 October 2026
Average block since genesis578 sTarget 600 s

Mainnet targets 600 seconds per block. Moving beyond 20 minutes does not trigger the testing chains' difficulty exception.

Live blocks come from mempool.space on 21 October 2026. Miners set timestamps, which can be ahead of real time.

Test coins, faucets and incentivized testnets

A testnet coin faucet distributes coins intended for testing, not payment. Access is not necessarily unrestricted, and a displayed wallet balance is not evidence of economic value. A reward program is also separate from the test coins themselves. Those distinctions help explain both access limits and misleading claims.

Faucet access requirements

In October 2026, faucets may require balances, accounts or browser proof of work. Alchemy caps Sepolia claims at 0.1 ETH daily, requiring at least 0.001 ETH on Ethereum mainnet. Coinbase requires an API key.

Prices despite policy

Goerli ETH traded at up to $1.60 on 25 February 2023. BIP-94 reported test bitcoin being actively bought and sold on testnet3 by 2024. Sales of testnet4 coins were reported in 2025.

Misleading payments and conversions

Scammers may pay in test ETH while a wallet shows a misleading fiat value. Offers to convert test bitcoin or signet coins into real BTC are scams.

Participation and Sybil activity

Celestia's 2023 Blockspace Race drew over 20,000 applicants for 1,000 places. Linea's 2023 test phase recorded 5.5 million unique wallets and over 47 million transactions. Human Passport described reward-seeking Sybils and bots in 2025.

What a testnet cannot show

Successful testnet execution is evidence about one environment, not a guarantee of mainnet behavior. Low economic stakes also leave some attacks weak or absent. A smart contract audit and state-based rehearsals address gaps that a public deployment alone cannot cover. These checks complement one another rather than providing interchangeable assurances.

Liquidity and oracle data

Isolated state lacks mainnet liquidity, live oracle prices and deployed dependencies. A mainnet fork reproduces existing state for interaction tests, rather than requiring those dependencies to be recreated.

Fees and network load

Cheap transactions do not establish production cost assumptions. Gas limits follow client testing plans, including Hoodi's scheduled rise to 200 million on 26 October 2026. Observed block timing is also environment-specific.

Configuration and deployment settings

A successful rehearsal does not validate the destination's settings. Deployment checks still need to account for contract addresses and configuration specific to each target.

Validators and MEV

A permissioned validator set does not reproduce open participation. Limited operator coverage also constrains what a rehearsal can show about MEV.

From rehearsal to production deployment

PixelPlex treats deployment as a production release, not just a successful testnet transaction. Its smart contract development process moves from rehearsal to verified bytecode and an operations-ready handover. The work spans a release-readiness checklist, mainnet bytecode verification and ownership transfer. The following parts organize that transition around evidence a delivery team can review, rather than a single passing result.

PixelPlex case icon

PixelPlex rehearses contract releases from testnet to mainnet. Its Canton wallet development work keeps participant nodes, synchronizers and ledgers in sync across LocalNet, DevNet, TestNet and production. PixelPlex also built Web3 Antivirus for a client. It runs user transactions on a dedicated testnet and analyzes the contracts behind them before users sign.

  1. Environment plan

    Define what each stage must demonstrate, from local behavior to testnet deployment. Record whether state persists, resets or comes from a fork. Include the intended chain ID.

  2. Deployment scripts and bytecode

    Keep deployment steps explicit and repeatable. Document the target configuration alongside them. Verified bytecode connects the rehearsed contract to the intended production release.

  3. Test data and access

    Account for faucet limits and the balances testing needs. Keep test data separate from assumptions about production liquidity or live prices. Document the applicable claim requirements.

  4. Monitoring during rehearsals

    Check block production, finality and transaction outcomes during QA and testing. Record anomalies with the network configuration that produced them. Compare observed behavior with the rehearsal's purpose.

  5. Production release checklist

    Confirm readiness before deployment, then prepare an operations-ready handover. The checklist should make configuration, verified bytecode and ownership transfer clear to the receiving team. Document that team's responsibilities.

Plan a release around your testing needs

We can discuss your target environment, deployment stages and handover requirements when your team is ready.

Discuss your project

Testnet FAQ

What is a testnet in crypto?

A testnet is a separate blockchain where developers rehearse applications and protocol changes before release. The term describes an environment, not an investment or a payment asset. Seeing the label in a wallet means the selected chain is intended for testing. A testnet label alone does not establish who operates the chain or whether a reward claim is genuine.

How does a testnet differ from mainnet?

Mainnet carries live assets, while a testnet maintains separate state and coins. Ethereum uses chain ID 1 for mainnet, 11155111 for Sepolia and 560048 for Hoodi. Those identifiers support replay protection under EIP-155, introduced in 2016. Bitcoin distinguishes its chains with network magic, ports and address prefixes rather than Ethereum's chain ID system. Bitcoin mainnet addresses begin with 1, 3 or bc1.

Which Ethereum testnet serves which purpose in 2026?

Sepolia serves Ethereum apps in October 2026, while Hoodi supports validator and staking work. Anyone can run a validator on Hoodi. Lifecycle dates also need context. Hoodi's published end dates differ, with the Ethereum Foundation listing September 2028 and the client README listing December 2028. A successor for application testing remains unnamed in October 2026. A May 2026 proposal discusses replacement during 2027.

How can developers get testnet ETH?

Developers get test ETH from faucets, which all impose caps. In October 2026, most Sepolia faucets require a small mainnet balance, an account or API key, or browser proof of work. Alchemy requires 0.001 ETH and Chainstack requires 0.08 ETH. With proof of work and a captcha, pk910 allows up to 2.5 Sepolia ETH or 33 Hoodi ETH per 24-hour session.

Do testnet coins have any real value?

Testnet coins are supposed to have no value, even when scarcity creates a secondary market. A quoted price does not change their intended role. Wallet displays can also assign misleading dollar values to test coins. Claims that test bitcoin can be withdrawn as real BTC are a documented scam pattern.

What is a devnet used to test?

A devnet is a short-lived network Ethereum client teams launch to test one upgrade before public testnets. Glamsterdam devnets 8, 9 and 11 ran in August and September 2026. Solana uses the name for application testing, with possible ledger resets. In October 2026, Canton resets DevNet about every three months and TestNet every three to six months, but MainNet never resets.

Can contracts pass testing but fail on mainnet?

A smart contract can pass a testnet rehearsal but fail with mainnet state, costs and incentives. Mainnet forks help test dependencies, but their controlled environment provides only one source of evidence. Deployment settings and economic assumptions remain separate checks, even after transactions execute successfully on a public testnet. Test coins cannot fully reproduce attacks driven by real economic stakes.

Related terms

All definitions
Sources (45)

Get updates about blockchain, technologies and our company

We will process the personal data you provide in accordance with our Privacy policy. You can unsubscribe or change your preferences at any time by clicking the link in any email.

Follow us on social networks and don't miss the latest tech news

  • facebook
  • X
  • linkedin
  • instagram
Stay tuned and add value to your feed