ChainSafe’s Lodestar v1.50.0 is a Hoodi clock, not an Ethereum mainnet fire. The consensus-layer client tagged the release on 12 October 2026 at 13:02 UTC as mandatory for Hoodi operators and only recommended for mainnet users. The same tag ships a Gloas (Glamsterdam) fork epoch on Hoodi and a 200 million gas-limit schedule. That is shipped software for a public testnet. It is not a live upgrade of Ethereum.
Ether sat near $2,499.61 on the desk snapshot at 2026-10-13T23:48:25+00:00 (CoinGecko via the site market plugin, 24-hour change +2.10%). Bitcoin was $83,185 on the same print. None of those prints depend on Lodestar 1.50.0. The fee tape is still a mainnet story. The fork clock in this tag is a Hoodi story.
I care about client releases because they are how upgrades actually arrive. I also refuse to treat every GitHub tag as a mainnet event. Show me the fee, and show me which network the epoch is bound to.
What happened
Lodestar v1.50.0 is a stable, non-prerelease tag. The notes tell operators to upgrade the Lodestar beacon node and validator client together. Hoodi operators must also move their execution client to a Glamsterdam-ready release before the fork. Building blocks works differently under Gloas, so a mismatched pair is the failure mode this tag is trying to prevent.
The schedule itself is in the merged config change, pull request 10313 (merged 8 October 2026 at 20:31 UTC). It sets GLOAS_FORK_EPOCH to 132352 on Hoodi: slot 4,235,264, Unix time 1793036568, which the PR states as Monday 26 October 2026, 17:42:48 UTC. The same epoch raises the gas limit to 200,000,000. Fork version 0x80000910 was already in the tree. Until this PR, the Hoodi Gloas epoch was still Infinity.
The release is explicit about validator plumbing after Gloas. Fee recipient and gas limit come from signed proposer preferences. From the Gloas fork onward, a Lodestar beacon node rejects prepare_beacon_proposer and register_validator with HTTP 410. If you run a different validator client against a Lodestar beacon node, that other client has to speak proposer preferences. Builder bids move in-protocol over p2p. External builders are configured on the validator client (--builder.urls, or per key), not as relays glued onto the beacon node. Pre-Gloas builder settings keep working until the fork. That is a real ops change. It is still gated on the Hoodi epoch, not on today’s mainnet slot.
Other items in the tag matter for operators and do not rewrite the network map: automatic use of the execution-layer REST/SSZ engine API when the execution client advertises it, a note that the notifyNewPayload timing metric will read lower because it no longer includes part of Lodestar’s own processing, and a pile of Gloas builder and proposer-preference features listed in the changelog. None of those lines schedule mainnet.
Context
The Ethereum Foundation’s 17 September 2026 Glamsterdam testnet post scheduled Sepolia at epoch 353,024 (6 October 2026, 13:53:36 UTC) and left Hoodi and mainnet as TBD. Lodestar 1.50.0 is the later client-side step that binds Hoodi. Sepolia was the first public Gloas-plus-Amsterdam run. Hoodi is the next public network in that sequence. Mainnet still has no epoch in this tag.
That sequence matches how Ethereum ships consensus changes under proof of stake: testnets first, then a mainnet epoch in a later coordinated release. We already treated Teku 26.10.0 the same way. See our Teku 26.10.0 read. Gloas headers and proposer-preference plumbing can land in a client without the chain under your feet forking today.
The live fee picture is still mainnet. DefiLlama’s Ethereum fees overview, read at 2026-10-14T00:13:05Z from api.llama.fi/overview/fees/ethereum, showed $10.49 million in 24-hour fees and $79.30 million over seven days. The daily chart in that payload prints $10.49 million for 12 October 2026 UTC and $10.81 million for 13 October 2026 UTC. Same pull of DefiLlama chain TVL listed Ethereum at $51.23 billion. Those are usage numbers on the chain people settle on. They are not a Hoodi metric, and they do not turn a recommended mainnet tag into a mandatory one. For the vocabulary, see gas fees.
Coinbase BTC-USD stats around the same hour (Chrome UA, 2026-10-14T00:12Z class pull) showed last $83,333.34 with a 24-hour range $81,329.58–$84,070.33. That tape is context for risk appetite. It is not evidence that Gloas has activated on mainnet.
Lodestar is one consensus client among several. A Hoodi-mandatory tag from ChainSafe does not, by itself, move Lighthouse, Teku, Nimbus, or Prysm. Other teams will publish their own Hoodi-ready notes. Until those notes exist, “the network is ready” is a slogan, not a checklist. Ethereum’s recorded history is full of upgrades that looked imminent in client changelogs and still waited on a coordinated mainnet epoch.
Our read
I read v1.50.0 as Hoodi scheduling plus a Gloas builder/validator API break, recommended on mainnet. I do not read it as Ethereum activating Glamsterdam, ePBS-style in-protocol builders, or a 200 million gas limit on mainnet this week.
The 410 responses and the move of builder URLs onto the validator client are the parts I would not skip if I ran Hoodi. They are exactly the class of “it compiled, but my proposer flow died at the fork” bugs. Mainnet operators can still install 1.50.0 because ChainSafe recommends it. They should not tell themselves the chain forked because a GitHub badge turned green.
Falsifiable claim: By 2026-10-26 23:59 UTC, Hoodi activates Gloas at epoch 132352 (slot 4,235,264, 17:42:48 UTC) as set in Lodestar PR 10313 and shipped in v1.50.0, and ChainSafe’s v1.50.0 notes still describe the tag as mandatory for Hoodi and only recommended for mainnet. If Hoodi misses that epoch without a replacement schedule in a later Lodestar tag, or if ChainSafe amends 1.50.0 to call the tag mainnet-mandatory or to drop the Hoodi epoch, this read is wrong.
What to watch next
Watch the Hoodi slot clock into 26 October 2026, 17:42:48 UTC. A clean Gloas activation at epoch 132352 with the 200 million gas-limit bump is the confirmation this tag promised. A skipped epoch, an emergency client, or a note that 10313’s timestamp was wrong would break the claim above.
Watch execution-client pairings on Hoodi. Lodestar is only the consensus side. The release tells Hoodi operators to be on a Glamsterdam-ready execution client before the fork. A Lodestar-only bump with a stale execution binary is how testnets look “upgraded” in a status page and still fail payload import.
Watch other consensus clients for Hoodi-ready stables, not blog rumors. Teku already shipped Gloas-header work. Lodestar now ships the Hoodi epoch. Lighthouse’s latest stable we see remains v8.2.3 (1 October 2026). A Hoodi date that exists in one client and not in the others is still a client story.
Watch mainnet fees, not the roadmap slide. If Ethereum fee revenue on DefiLlama’s Ethereum overview collapses because of a client bug after people rush 1.50.0 onto mainnet, that would be a different article. Today’s print is still in the $10 million 24-hour neighborhood. For a plain method on timestamped prints, see how to read a market snapshot. More Ethereum desk copy lives under Ethereum news.