Spent the last few days digging into OpenGradient, mostly because "verifiable AI" gets thrown around constantly without much substance behind it.
The core idea is actually pretty simple once you strip the jargon away: every time an AI model runs on their network, it comes with a kind of receipt — proof of which model ran, what input it got, and what it spit out. You don't have to take anyone's word for it.
What caught my attention was the funding. They raised $9.5 million, backed by a16z crypto and Coinbase Ventures, which isn't huge by crypto standards, but it's a meaningful vote of confidence in a space that's mostly noise right now.
The token launched in April 2026 on Binance, and adoption numbers since then aren't nothing — over 2 million verifiable inferences processed, more than 2,000 AI models hosted on their Model Hub, and north of 500,000 cryptographic proofs generated. That's actually pretty interesting because most "AI x crypto" projects show you a roadmap PDF, not usage data.
Most projects talk about decentralizing AI. The difference here is they built something closer to a permissionless Hugging Face, where anyone can upload a model and have it instantly queryable across the network.
Still, a chunk of current activity feels tied to the Binance listing rather than organic demand. Mainnet and real fee utility are still ahead.
Still early, but something seems to be taking shape here.
#OPG @OpenGradient $OPG
$HEI $SAHARA
The core idea is actually pretty simple once you strip the jargon away: every time an AI model runs on their network, it comes with a kind of receipt — proof of which model ran, what input it got, and what it spit out. You don't have to take anyone's word for it.
What caught my attention was the funding. They raised $9.5 million, backed by a16z crypto and Coinbase Ventures, which isn't huge by crypto standards, but it's a meaningful vote of confidence in a space that's mostly noise right now.
The token launched in April 2026 on Binance, and adoption numbers since then aren't nothing — over 2 million verifiable inferences processed, more than 2,000 AI models hosted on their Model Hub, and north of 500,000 cryptographic proofs generated. That's actually pretty interesting because most "AI x crypto" projects show you a roadmap PDF, not usage data.
Most projects talk about decentralizing AI. The difference here is they built something closer to a permissionless Hugging Face, where anyone can upload a model and have it instantly queryable across the network.
Still, a chunk of current activity feels tied to the Binance listing rather than organic demand. Mainnet and real fee utility are still ahead.
Still early, but something seems to be taking shape here.
#OPG @OpenGradient $OPG
$HEI $SAHARA