BREAKING: 🇺🇸 President Trump is set to make a “huge” announcement today at 5:00 PM ET.
Sources are speculating it could involve plans to reopen the Strait of Hormuz and a possible new peace deal with Iran. Markets could see major volatility if confirmed.
BREAKING: The ceasefire is already under pressure.
The 🇺🇸 U.S. has struck Iranian missile sites, drone depots, and radar systems after Iran reportedly attacked a ship exiting the Strait of Hormuz marking the first U.S. strike on Iran since the MOU was signed.
$AAVE looks stronger than BTC here. It reclaimed the key zone around $85.45, and if it holds, the next upside targets are clean. Current AAVE is around $86.1.
Been going deeper on @OpenGradient this week and one architectural detail keeps pulling my attention.
OpenGradient never actually puts model weights or zkML proof payloads on-chain. It posts a Blob ID. That's the whole trick. Heavy data lives on Walrus, the 32-byte pointer lives on the chain and that separation is likely why the network has been sustaining 10,000+ daily transactions across 263,500+ unique wallets without visible throughput degradation.
#OPG What made me sit with this longer was the Upbit listing on June 15. Volume hit roughly $358M in a single session a 606% spike which meant a wave of first-time wallets suddenly touching $OPG
Some portion of those almost certainly explored the Model Hub. If each inference request pulled full model binaries through the chain, a demand event that size would be genuinely destabilizing. The Blob ID architecture is essentially load insulation and it held.
I'll be honest, I don't fully understand how inference node caching behaves under that kind of sudden influx. Nodes download models using the Blob ID if not already cached (Opengradient) so what happens when dozens of new nodes need the same model simultaneously? Does Walrus handle concurrent shard pulls gracefully at that scale?
The separation of storage and settlement is clean design. What I'm less sure about is whether the retrieval layer was actually tested by June 15 or whether that volume stayed at the exchange layer and never touched inference demand at all.
@OpenGradient has been sitting in my open tabs all week. The thing that actually stopped me wasn't the price it was watching the TEE inference count keep climbing while the token bled.
OPG dropped ~12% in 24h and is now trading around $0.154, with $40.7M in daily volume a significant retreat from the $357M spike that the Upbit listing caused on June 15. Classic post-listing behavior, nothing unusual there. But the compute side didn't seem to care. The 150,000+ privately executed TEE inferences kept accumulating which tells you something about who's actually in the network.
Developers querying the Model Hub don't stop building because Korean retail momentum faded. That split price action decoupling from throughput activity is usually more telling than the volume spike itself.
The broader network stats back this up: 4.2M+ blocks produced, 1.85M+ on-chain transactions, 10,000+ daily transactions, and 263,500+ unique wallets interacting with the system. But wallet count is a headline number.
It says nothing about repeat usage depth or whether those wallets are active agents versus dormant holders. What I genuinely can't resolve yet is whether the TEE inference volume is fee-paying compute or pre-production testing.
Every verified call is supposed to settle in OPG on Base in real time (Opengradient) but mainnet utility at full scale is still ahead. So the real question: is the inference demand we're watching actually flowing through the x402 payment layer, or just warming up?