Obol (OBOL): distributed validators for Ethereum staking, and what makes it risky

Obol Network builds software that lets several independent operators run one Ethereum validator together, so that no single machine holds the whole signing key. On 9 November 2026 DefiLlama counted about $1.21 billion of ETH staked on Obol distributed validators, up from $1.11 billion a week earlier, at a time when most large staking protocols lost value. The OBOL token tells a different story: CoinGecko showed a market capitalisation of about $780,000 at 19:33 UTC the same day, and the token carries no governance rights under its own MiCA white paper.

That gap between the stake running on the technology and the value of the token is the main thing to understand about Obol crypto. This page explains how the Charon client and distributed key generation work, who builds and controls the project, what OBOL holders get, where the recent growth came from, and which risks remain, with every figure dated.

What is Obol Network? Distributed validators for Ethereum staking

An Ethereum validator signs attestations and proposes blocks with a single BLS key. In the usual setup that key sits on one server run by one operator. If the server goes offline, the validator misses rewards; if the key leaks, the stake behind it can be slashed. Distributed validator technology (DVT) splits that job across a cluster of machines, each run by a different operator, and asks a quorum of them to agree before anything is signed.

Along with SSV Network, Obol is one of the two main DVT providers on Ethereum. Its brand covers several things. Obol Labs, the original company, was renamed DV Labs on 19 August 2024 and writes most of the code. The Obol Association, a non-profit association registered in Zug, Switzerland, on 26 July 2023, issued the OBOL token and runs the treasury and grants. The Obol Collective is the wider group of staking protocols, client teams, tool makers and node operators that build on or around the software, funded in part by retroactive grants.

DefiLlama lists Obol as a staking protocol on Ethereum only, with the methodology “Total value of ETH staked on Obol’s Distributed Validators”. Obol does not issue a liquid staking token of its own. Stakers usually reach it through another protocol, above all Lido, whose node operators run part of their validators on Obol clusters.

How a distributed validator signs: Charon, DKG and threshold keys

The core program is Charon, a middleware written in Go. Each operator in a cluster keeps its own consensus client and validator client, and puts Charon between the two. The validator client thinks it is talking to a normal beacon node; in fact Charon passes the duty to the other nodes of the cluster, collects partial signatures and combines them once enough have arrived.

The key itself is created by a distributed key generation ceremony (DKG), run through the DV Launchpad web interface. At the end of the ceremony each node holds a key share, and the full private key is never assembled in one place. A signature becomes valid when a threshold of shares has signed. Obol’s examples use clusters of four nodes with a threshold of three, and clusters of seven with a threshold of five, where two nodes can fail without the validator going dark. The MiCA white paper describes the general rule as tolerating up to one third of the nodes failing.

The design changes the risk, it does not remove it. A four-node cluster with a 3-of-4 threshold survives one outage, not two. All the nodes also have to talk to each other within a slot: in the Simple DVT trials run with Lido, latency above 235 milliseconds inside a cluster hurt performance, according to Obol. Slashing protection still matters, because a cluster that signs two conflicting messages is punished like any other validator. Our guide to validators, slashing and rewards covers those basics.

Client diversity is the other selling point. Each operator can run a different execution and consensus client, which limits the damage of a single client bug. A second DV client is arriving. Pluto, a Rust rewrite of Charon by Nethermind, has had a public repository since 11 March 2026. In August 2026 a four-node cluster on the Hoodi testnet ran two Charon and two Pluto nodes side by side, and Obol said some early mainnet deployments were under way. Charon remains the client almost all Obol clusters use.

The tools around Charon

Beyond Charon and the Launchpad, the Obol stack includes Obol Splits, a set of smart contracts that divide execution and consensus rewards between the operators of a cluster; an SDK and API for staking protocols; Techne credentials, an on-chain record of operator skills; and monitoring tools such as the Fleet-to-Node dashboard launched on 22 January 2026. On 29 October 2026 Obol published ValidatorBeat, an in-browser tool that scores how exposed a validator setup is to correlated failures, with the support of Ethereum Institutional. Answers stay in the browser and are not published.

The code lives in the ObolNetwork organisation on GitHub, created in June 2021 and holding 72 public repositories on 9 November 2026. The latest Charon release, v1.11.4, was published on 9 November 2026 at 09:17 UTC as a recommended patch. It makes DKG ceremonies more reliable and redacts secret values from the client command lines that its diagnostics export. Its notes say plainly that it is not ready for Gloas, the consensus side of the next Ethereum upgrade, Glamsterdam: support is planned for Charon v2.0.0, which existed only as release candidates (rc1 on 20 October, rc2 on 21 October) on that date.

Who builds Obol and who decides

Collin Myers and Oisín Kyne founded Obol Labs, according to the DV Labs launch post of 19 August 2024, which dates the start to more than three years earlier; the GitHub organisation was created in June 2021. DV Labs describes itself as a remote development team of more than 30 people in more than 15 countries. It builds Charon, the Launchpad, the SDK and the API.

The token side belongs to the Obol Association, at Baarerstrasse 10 in Zug. Its MiCA white paper names two members of the management body: Francis Hackett, President of the Board, and Patrick Storchenegger, Board Member. The association sold the token, holds the treasury and pays grants. In July 2025 it announced that it would mint the remaining supply up to the 500 million cap, saying it did so for transparency; the white paper and the forum post do not say which address or multisig controls minting.

Governance is advisory. OBOL holders can delegate voting power to delegates in a “Token House”, but the votes are non-binding, and the white paper states that the final implementation of any proposal “remains at the discretion of the Obol Association”. The official documentation does not describe a timelock or multisig for the Obol Splits contracts or the OBOL staking contract. For a reader, that means trust in Obol is trust in a small Swiss association and a development company, not in token holders.

Lido, Simple DVT and the move to 0x02 validators

Lido is the main reason Obol matters. In spring 2024 Lido opened its Simple DVT module, in which clusters of seven operators (Obol or SSV) each ran validators with a 5-of-7 threshold. At its one-year mark, Obol says, the module held 308,320 ETH, 3.39% of Lido deposits and 0.88% of all staked ETH.

Lido then changed course. The 72 regular Simple DVT clusters were wound down after a Snapshot vote, effective 1 July 2026 according to Obol, and Lido’s own blog of 27 July 2026 confirms they “have been wound down”. The larger Super Clusters, about 160,000 ETH by Obol’s count, were not affected. Small operators can now join through Identified DVT Clusters in Lido’s Community Staking Module, run on Obol or SSV.

The bigger shift came with Lido’s Curated Module v2, presented on Lido’s blog on 27 July 2026. It lets more than 265,000 curated validators move from the old 0x01 withdrawal credentials to 0x02 “compounding” credentials, introduced by the Pectra upgrade, which raise a validator’s maximum effective balance from 32 ETH to 2,048 ETH. Lido warned that, with an activation queue of more than 40 days, the migration will take time. Several Lido curated operators have said they are putting part of that stake on Obol clusters: RockLogic on 3 August, Blockdaemon on 13 August and DSRV on 15 September 2026, which set itself a target of 20% of about 250,000 ETH. Stakely, ParaFi Tech, Pier Two and GatewayFM appear in Obol posts too, and an Obol post of 2 July 2026 says Bitcoin Suisse has moved its Ethereum staking to Obol distributed validators. Our news piece on Lido’s switch to larger validators covers that migration.

Larger keys raise the stakes of each cluster. Obol itself points out that a 2,048 ETH key costs 64 times more per minute of downtime than a 32 ETH key, which is the argument for spreading such keys across several operators in the first place.

Other partners are older or smaller. ether.fi committed in April 2024, according to Obol, to deploy more than $1 billion of stake on Obol distributed validators through seven operators; no current figure for ether.fi stake on Obol has been published. Obol was a day-one partner when Lido V3 and its stVaults launched in the first quarter of 2026, and it announced a partnership with Primev, the mev-commit project, on 3 March 2026.

Obol TVL rose about 9% in a week while the sector fell

DefiLlama recorded $1.110 billion of ETH on Obol distributed validators on 2 November 2026 and $1.212 billion at 18:56 UTC on 9 November, a rise of about 9% in dollars. Measured in ETH, the change is larger: from 489,254 ETH to 570,205 ETH, about 16.5%, because the ether price fell over the week. Over the same days, DefiLlama shows Lido falling from $22.05 billion to $20.64 billion, ether.fi from $4.77 billion to $4.46 billion and SSV Network from $11.88 billion to $11.12 billion, all down about 6%.

Obol’s public validator data shows how the rise happened. The number of validator keys barely moved, from 6,286 to 6,298. The 17 new keys carried 804 ETH. Most of the increase came from validators that already existed and grew: 185 of them gained more than 1 ETH each, adding 80,235 ETH in total. Keys holding more than 33 ETH went from 194 to 348, and typical cases jumped from about 32 ETH to well over 1,000 ETH.

That pattern is consistent with consolidation into 0x02 validators, where an operator merges many 32 ETH validators into a few large ones, as Lido’s curated operators have been doing under CMv2. It is not an established fact. Neither Obol nor Lido has said which operators moved this stake onto Obol clusters during the week, and the public data does not link keys to operators. ETH consolidated into an Obol key may also have been staked with another setup before, so the figure says more about where stake sits than about new money entering Ethereum staking.

One technical caveat: DefiLlama uses Obol’s daily snapshots for history and live beacon-chain balances for its latest point, so small gaps between two readings on the same day, such as $1.223 billion and $1.212 billion on 9 November, are differences of method. The all-time high in DefiLlama’s series is about $3.05 billion, on 23 August 2025.

The OBOL token: no governance rights, a small market value

OBOL is an ERC-20 token on Ethereum (contract 0x0b010000b7624eb9b3dfbc279673c76e9d29d5f7) with a fixed supply of 500 million. CoinGecko showed about 342.9 million in circulation on 9 November 2026. The token went on sale on CoinList from 24 February to 3 March 2025 at $0.25, an implied fully diluted value of $125 million, and trading started on 7 May 2025. Earlier investors came in at fully diluted values of $50 million (seed, 2021), $125 million (Series A, 2022) and $180 million (strategic round, 2023), according to the white paper, which dates the Series A to 2022 although it was announced in January 2023.

At 19:33 UTC on 9 November 2026, CoinGecko put OBOL’s market capitalisation near $780,000 and its fully diluted value near $1.14 million, with the token trading below one cent. CoinGecko lists an all-time high of $0.3801. On those readings, the token’s market value was below 0.1% of the ETH staked on Obol clusters. That says nothing about a fault; it does show that the stake running on Obol does not flow to OBOL holders.

The white paper is clear on rights. It says the OBOL token “has no rights or obligations within the Obol Collective” and does not give governance powers or enforceable claims. What holders get is a say in advisory votes, a role in the retroactive funding programme and, in plans written in 2025, staking and restaking options. The same document planned a capped reward programme of 2.5% of supply a year for ETH staked through distributed validators. The white paper was notified under MiCA with France as home member state; it states that it “has not been approved by any competent authority” in the EU, and that the token may lose its value in part or in full.

Allocation and unlocks

Obol’s documentation gives this split of the 500 million tokens: Ecosystem Treasury and retroactive funding 38.8%, investors 23.7%, team 19%, community incentives 7.5%, airdrop 7.5%, and the CoinList public sale 3.6% (the published figures add up to 100.1%). The documentation does not detail vesting. Tokenomics.com, a third-party tracker, shows the investor and team cliffs ending on 7 May 2026, team tokens released quarterly until 7 May 2028, and the treasury vesting until August 2028. It lists the next token unlock on 7 December 2026: 4,956,482 OBOL, about 1% of supply, for community incentives.

Buybacks into liquidity, not burns

In January 2026 Obol used part of the ETH earned from distributed validator fees to buy 1,706,309 OBOL, which it placed in Arrakis-managed Uniswap vaults as protocol-owned liquidity. Obol’s first-quarter report says a second operation of the same kind followed during the quarter. The posts describe no burns and no payouts to holders, and no fee rate is published. The same report says Obol halted trading on “a handful of centralized exchanges” to concentrate liquidity, without naming them. The OBOL staking contract tracked by DefiLlama held only about $5,600 on 9 November 2026.

Fees and revenue: small numbers for now

DefiLlama counts the fees generated by Obol services at $52,985 over the 30 days to 9 November 2026 and $545,841 since tracking began. The share collected by Obol itself, which DefiLlama labels revenue, was $2,551 over those 30 days and $72,571 in total. Obol’s posts do not give the fee rate behind that gap. Fees paid inside Lido’s modules, such as the 2% or 4% DAO fees in Simple DVT, go to Lido, not to Obol.

For a protocol securing more than $1 billion, that is a thin revenue base. It also limits how far treasury buybacks can go without spending the treasury itself.

Audits and the security record

Obol keeps its audit reports in the obol-security repository on GitHub. They are:

• Sigma Prime, June 2023, on the Charon client (report version 2.1).
• Zach Obront, 18 to 22 September 2023, on the Obol Manager contracts that split rewards and route withdrawals, including the Lido split contracts.
• Quantstamp, review from 20 November to 19 December 2023, final report on 2 February 2024, on Charon: 18 findings, including three high-severity issues, all three fixed.
• Sayfer, March 2024, a penetration test of the DV Launchpad.
• Trail of Bits, kick-off on 26 January 2026 and final report on 20 February 2026, on the Pedersen DKG and the reshare logic behind Charon’s edit commands: nine findings (six medium, two low, one informational, none high), all fixed in Charon v1.9.0, according to Obol.

The older reviews covered versions of Charon long since replaced; the code now runs at v1.11 with v2.0 in preparation. Obol has published no audit of the OBOL token or of the OBOL staking contract. DefiLlama’s hack database lists no incident for Obol, and no loss of staked funds has been reported.

The most recent alarm came from outside. In October 2026 a certificate authority problem at eMudhra, an Indian CA, led Obol to ship Charon v1.11.2 on 21 October as a release required for all operators. No source reported affected validators or funds. Our article on the eMudhra weak-keys case has the details.

Obol vs SSV Network and other choices

SSV Network is the direct rival. Both offer threshold-signed validators, and Lido accepts both in its DVT modules. On DefiLlama, SSV held about $11.1 billion on 9 November 2026, roughly nine times Obol’s total. The two differ in design: Obol’s Charon sits as middleware next to existing clients, and clusters are formed by operators who choose each other through a DKG ceremony.

For a staker, the realistic choice is rarely Obol alone. It is a staking provider, a liquid staking protocol or a vault that may or may not use Obol underneath. Solo stakers can run a cluster with friends through the Launchpad, but they still need hardware, uptime and a working relationship with the other operators.

Why Coindaro rates Obol risk 3 out of 5

On the positive side, the software has been audited several times, including in 2026, it runs live stake for Lido operators and others, and no loss of funds has been recorded. The design reduces single-operator risk, which is the problem it was built for.

On the other side, the token gives no rights and its market value is tiny compared with the stake on Obol; revenue retained by Obol is small; decisions rest with the Obol Association and DV Labs; growth depends heavily on Lido’s choices, which already closed the regular Simple DVT clusters once; and Charon still has to ship a stable v2.0.0 for Glamsterdam. The middle score reflects a working technology wrapped around a weak token case.

Dates to watch

The next scheduled OBOL unlock is on 7 December 2026, according to Tokenomics.com. Lido targets the fourth quarter of 2026 for a dedicated module that lets permissionless node operators use 0x02 credentials, which could bring more large validators to DVT clusters. The mainnet date of Glamsterdam, and with it the release of Charon v2.0.0 for production, had not been fixed on 9 November 2026. Obol has also recommended that 30% of Lido’s curated validators run on its distributed validators by the end of 2027: a recommendation, not a commitment by Lido.