#opg $OPG I recently took a careful look through OpenGradient’s full set of materials—from the official website, to development documentation, to open-source repositories, and even Base chain token holding data. It’s clear that this blockchain’s AI project has an extremely sharp contrast between strengths and weaknesses.
The project’s foundation is actually quite solid. Backed by NVIDIA’s Inception incubation program, and with investment participation from multiple institutions including Coinbase Ventures. Well-known professionals in the industry have also publicly endorsed it. The core technology follows a ZKML combined with a TEE dual-verification approach, directly targeting the “black box” pain point of traditional AI that can’t verify the computation process. The GitHub repository comes with a complete set of tools: an open-source Python development SDK and the quant trading framework BitQuant. The testnet runs stably, having already completed over 2.0 million verifiable AI inference tasks. It also has more than 4,000 online model variants—real-world deployment of underlying development tools, not a project that just tells a story based on concepts. @OpenGradient
However, there are significant concerns on the tokenomics side. The total supply of OPG is 1 billion tokens, with highly concentrated holdings. The top ten wallets hold nearly 95% of the circulating supply. Large whales control the vast majority of the tokens, leaving ordinary retail holders with no bargaining power. Both the team’s and investors’ allocations come with a one-year lock-up period; after that, they are released linearly over multiple years. This means long-term continuous sell pressure is hard to avoid. With the on-chain AI sector currently booming, the price might be supported short-term by narratives or hype, but within the ecosystem there are very few real token consumption scenarios. That makes it difficult to absorb an ongoing, stream-like release of unlocked tokens.
Overall, the project has innovative highlights in its technical architecture and solid development support. But the token distribution mechanism has obvious loopholes. At this stage, entering the project is essentially the same as helping large holders raise the price. I plan to put it on my watchlist for ongoing observation first. Once the token supply is sufficiently dispersed and market liquidity improves, I’ll reassess whether there’s an opportunity to participate again.
The project’s foundation is actually quite solid. Backed by NVIDIA’s Inception incubation program, and with investment participation from multiple institutions including Coinbase Ventures. Well-known professionals in the industry have also publicly endorsed it. The core technology follows a ZKML combined with a TEE dual-verification approach, directly targeting the “black box” pain point of traditional AI that can’t verify the computation process. The GitHub repository comes with a complete set of tools: an open-source Python development SDK and the quant trading framework BitQuant. The testnet runs stably, having already completed over 2.0 million verifiable AI inference tasks. It also has more than 4,000 online model variants—real-world deployment of underlying development tools, not a project that just tells a story based on concepts. @OpenGradient
However, there are significant concerns on the tokenomics side. The total supply of OPG is 1 billion tokens, with highly concentrated holdings. The top ten wallets hold nearly 95% of the circulating supply. Large whales control the vast majority of the tokens, leaving ordinary retail holders with no bargaining power. Both the team’s and investors’ allocations come with a one-year lock-up period; after that, they are released linearly over multiple years. This means long-term continuous sell pressure is hard to avoid. With the on-chain AI sector currently booming, the price might be supported short-term by narratives or hype, but within the ecosystem there are very few real token consumption scenarios. That makes it difficult to absorb an ongoing, stream-like release of unlocked tokens.
Overall, the project has innovative highlights in its technical architecture and solid development support. But the token distribution mechanism has obvious loopholes. At this stage, entering the project is essentially the same as helping large holders raise the price. I plan to put it on my watchlist for ongoing observation first. Once the token supply is sufficiently dispersed and market liquidity improves, I’ll reassess whether there’s an opportunity to participate again.