Home » Blockchain » Base Denim Upgrade Goes Live, Cutting Block Times to 200ms

Base Denim Upgrade Goes Live, Cutting Block Times to 200ms

Within hours of its biggest speed upgrade yet, Base started flagging emptier-than-expected blocks. The fix for one blockchain’s bug has to outrun timers now ticking ten times faster than before.

Abstract blue streak of light representing faster blocks after the Base Denim upgrade

“Denim has successfully activated.” (Base status page, November 5, 2026, 18:14 UTC)

That one-line update is what Base, Coinbase’s Ethereum layer-2 network, posted to its own status page about 15 minutes after flipping the switch on the Base Denim upgrade. The change itself is simple to describe. Blocks that used to close every two seconds now close every 200 milliseconds, ten times more often. What’s not simple is what happened right after. In the same breath, Base said it was already investigating a new problem: more empty blocks than expected.

Denim isn’t really about raw speed, even though speed is the headline number. It’s about retiring a 16-month-old workaround and making the fast version of Base the only version. Here’s what changed, why developers wanted it, and what’s still shaky a day into the rollout.

What the Base Denim Upgrade Changes

Base activated Denim on mainnet at 18:00 UTC on November 5, 2026. The first 200-millisecond block, number 53,556,127, landed one second later, according to the crypto news outlet Unchained, which tracked the activation live. Base’s own documentation describes the shift in plain terms: block production moved “from one canonical block every two seconds to five complete canonical blocks per second.”

Every one of those five blocks gets its own hash, state root, and set of receipts, the same bookkeeping a block has always had, just five times a second instead of once every two seconds. Denim also adds a new onchain millisecond clock called BaseTime, and RPC responses now report timestamps down to the millisecond instead of the second.

Measure Before Denim After Denim
Block time 2 seconds 200 milliseconds
Canonical blocks per second 0.5 5
Fast data source for apps Flashblocks preview stream Canonical block and RPC streams
Node software required Any supported version base/base v1.5.1 or later

Source: Base upgrade documentation (docs.base.org) and Unchained, Nov. 5, 2026.

That last row matters more than it looks. Node operators who hadn’t updated to base/base v1.5.1 or later risked falling out of sync the moment Denim flipped on. A separate point release, v1.5.3, patched an unrelated state root bug discovered around the same time.

Why Retiring Flashblocks Is the Real Upgrade

To understand why Base bothered, you have to know what it’s replacing. Since July 2025, Base ran a system called Flashblocks, built with Flashbots, layered on top of its old two-second blocks. Flashblocks pushed out a partial, unofficial preview of each block every 200 milliseconds while the real, final block was still two seconds away from closing. Wallets and trading apps could show users something that felt instant, but that preview was never quite the truth. It could still change before the official block locked it in.

Denim deletes that gap. There’s no more preview layer sitting on top of a slower official record, because the fast version is now the official version. Base’s own migration guide is blunt about the tradeoff: “the ‘pending’ block tag and Flashblocks preconfirmation state have no direct equivalent after Denim.” Any app that was reading Flashblocks data has to switch to reading plain canonical blocks instead.

For everyday users, the difference “should feel small, since Flashblocks already offered near-instant confirmations,” Bankless writer William M. Peaster noted before the upgrade went live.

That’s the honest framing. This isn’t a speed upgrade people will feel at checkout. It’s a plumbing upgrade for the developers who had been building trading platforms, payment tools, and games around a preview feed that was never officially real. Base’s Cobalt upgrade this September already laid some of that groundwork by adding conditional transactions to the network. Denim is the second big mainnet change on Base this fall, and arguably the riskier one, since it touches how every single block on the chain gets made.

What We Don’t Know Yet

  • The empty-block problem isn’t fixed. Base’s status page says the team is “investigating periodic transaction inclusion delays, causing a higher than expected rate of empty blocks.” Unchained counted 146 of the first 300 post-Denim blocks carrying nothing but system transactions. As of the last update we could find, the status still read “Verifying,” not “Resolved.”
  • Base’s own documentation hadn’t caught up. Hours after the status page confirmed Denim was live, Base’s overview and migration pages still listed mainnet as “Planning.” A small sign the rollout outran Base’s own paperwork.
  • One app chain couldn’t keep pace. Horizen’s OP Stack layer-3 network, built on top of Base, stopped taking user transactions about 40 minutes after the switch flipped, because it couldn’t follow the new block timing.
  • Block-based timers now run roughly ten times faster. Any smart contract that counts a cooldown, a vesting schedule, or a governance voting window in block numbers rather than real time is affected. Unchained’s math shows a timelock sized for one week under the old 2-second blocks would lapse in under 17 hours now. Contracts that can’t be upgraded are stuck with whatever schedule they were given.

Who Benefits and How

Denim is infrastructure, not a feature shoppers will notice this week, so the benefit depends on who’s asking.

  • Developers building trading apps, payment tools, and games on Base no longer have to design around two competing versions of “what happened,” a fast preview and a slower official record. That should mean fewer edge-case bugs for any team willing to rebuild on canonical block and RPC streams instead of the retired Flashblocks feed.
  • People using apps already built on Base, including Shopify’s USDC checkout, likely won’t notice a difference yet. Flashblocks already made things feel close to instant, so the real gains show up only once those apps are rebuilt around guaranteed-final 200-millisecond blocks, and nobody has set a date for that.
  • Anyone running a Base node has to act now, not later: skipping the upgrade to base/base v1.5.1 or newer risks falling out of sync with the rest of the network.
  • Teams with contracts that measure time in block numbers, vesting cliffs, cooldown periods, voting windows, need to recheck that math today. A schedule built for 2-second blocks is quietly running about ten times faster than its authors planned.

Frequently Asked Questions

Is Base down right now?

No. Base confirmed Denim “successfully activated” and the network kept producing blocks throughout. The open issue is a higher than expected rate of empty blocks, which the team was still investigating as of its last public status update, not a network outage.

Do I need to do anything as a regular Base user?

No. Wallets and major apps handle the switch behind the scenes.

What was Flashblocks, in plain terms?

A preview system Base ran since July 2025, showing apps a likely version of a block up to two seconds before it was actually final. Denim makes that preview unnecessary.

Base has shipped the speed number it promised. Whether Denim counts as a clean win depends on something harder to measure: how long it takes Base’s engineers to explain why so many of the first blocks after launch came up empty. Watch the status page, not the announcement, before taking “it just works” at face value.

Sources

Related Reading