Argentum AI is deploying new capital from an oversubscribed round led by Kraken to build a decentralized network that lets anyone monetize idle processing power, creating a new class of “GPU entrepreneurs.”
Summary
- Kraken led an oversubscribed pre-seed round in Argentum AI, a Menlo Park startup building a decentralized compute marketplace.
- Argentum AI enables organizations to bid for computing power on-chain, letting providers monetize idle GPUs and data center capacity.
- The platform supports AI training, 3D rendering, scientific simulations, and more, blending human oversight with AI-driven task matching.
According to a press release shared with crypto.news on Oct. 1, the Menlo Park-based startup secured an oversubscribed pre-seed round. The capital infusion was led by crypto exchange Kraken, with participation from Banyan Ventures and angel investors Victor Morganstern and Todd Bensen.
Argentum AI said it will use the funds to accelerate the development of its core platform, an open marketplace that uses blockchain for real-time bidding and settlement of computational tasks.
You might also like: Crypto miners flock to Brazil on major clean energy surplus
Inside Argentum AI’s marketplace model
Argentum AI describes itself as a “human-friendly, AI-powered compute marketplace,” and its model blends decentralized infrastructure with machine intelligence to improve how computational jobs are matched, priced, and executed.
Per the release, Argentum AI functions as an open exchange where organizations can post computing tasks, ranging from AI model training and 3D rendering to big data analysis and digital twin simulations. Providers, whether individuals with a spare GPU or data centers with large clusters, bid to execute those tasks. Settlement occurs on-chain, with Ethereum smart contracts holding funds in escrow until successful completion
Rather than positioning AI as a replacement for human input, the platform emphasizes collaboration. Clients define requirements and oversee decision-making, while an embedded AI assistant learns from completed tasks to recommend better resource matches and pricing strategies over time. The result, according to the company, is a smarter and faster marketplace that adapts as it grows.
This model is built to tackle a specific and growing bottleneck in the tech world. According to CEO Andrew Sobko, the constraint is not just a lack of raw computing power, but a lack of flexible partners that can accommodate the unpredictable and varied demands of modern AI workloads. Argentum is engineered to solve that rigidity. Sobko further elaborated on the company’s phased approach to the market.
“In the immediate term, we see a massive opportunity for bridging flexible needs with stranded compute capacity. Longer term, AAI plans to work with GPU makers on establishing liquidity and monetization plans for second-life assets, further reducing the overall compute cost for customers,” Sobko said.
Target clients and tokenomics
According to the press release, Argentum AI’s initial target clients are industries with massive and fluctuating computational needs, including fintech companies, banks, and digital gaming companies. By allowing enterprises to source capacity on demand without committing to a single vendor, the project is positioning itself as a neutral layer in a market dominated by large cloud incumbents.
To fuel its internal economy, Argentum AI has architected a native SPL token, AGP (Argentum AI Point). The AGP token will serve as the native currency for computational services within the marketplace. Clients pay for workloads in AGP, while providers earn tokens by supplying capacity.
The model includes staking mechanisms that allow participants to boost visibility, unlock discounts, or access premium features. Governance rights are also built in, giving token holders a say in protocol upgrades and policy decisions. With a fixed supply of one billion AGP, the design leans on scarcity to align long-term incentives.
Read more: BNB Chain hack update: X account reclaimed after $13k theft, victims to be compensated
