Solana Cuts Slot Time to 250ms, Sets Sights on 200ms Target

TLDR Solana’s target slot time dropped to 250 milliseconds on September 18, 2026, at epoch 1037. Early data from a 60-minute sample showed slots landing at about 266ms, with a 0.05% skip rate. A draft proposal would cut per-slot compute limits as the network moves toward a possible 200ms slot time. Leader handoff windows shrink from one second to 800 milliseconds under the proposed 200ms setting. An August routing failure at hosting provider TeraSwitch knocked out reachability for 28.83% of network stake for about 33 minutes. Solana’s network now runs on a 250-millisecond target slot time. A slot is the window a validator gets to produce a block, so the shorter interval gives users more frequent chances for transactions to land. The change took effect at epoch 1037 on September 18, 2026, according to the Solana engineering changelog.Solana Compass placed the exact transition around 05:06 UTC that day. A sample taken on September 20 covered 60 one-minute windows. It found slots landing at roughly 266 milliseconds on average. Epoch 1037 skipped about 0.05% of its scheduled slots. A Tighter Compute Budget A draft design called SIMD-0525 would take the next step down to 200 milliseconds. That proposal also lowers the per-slot compute budget. At 250ms, the budget sits at 62.5 million compute units. At 200ms, it would drop to 50 million. The network’s overall ceiling stays near 250 million compute units per second in both cases. This means blocks would arrive more often but each would carry less permitted work. Actual transaction throughput still depends on demand and how leaders fill available blockspace. Handoffs, Geography and Infrastructure The four-slot leader window stays fixed under the new design. That gives each leader one second at 250ms and 800 milliseconds at 200ms to receive traffic and produce blocks. Geography affects that window. A Solana Foundation analysis found a median first-slot delay of about 28 milliseconds when consecutive leaders sat less than 500 kilometers apart. That delay grew to about 122 milliseconds when leaders were more than 8,000 kilometers apart. That figure equals 61% of a 200ms target slot. Engineers are testing two fixes. Agave developers are building forwarding methods that send transactions ahead when they might miss their intended leader. Client teams are also testing block and transaction execution against shared conformance standards. The repair-defer threshold, meanwhile, stays at 250ms even under the 200ms proposal. A separate event in August showed how infrastructure concentration can affect the network. On August 12, a routing failure at hosting provider TeraSwitch caused 12 sites to lose reachability. Solana Compass measured 28.83% of network stake as delinquent for about 33 minutes during that incident. The Solana Foundation said blocks kept landing throughout.