Brett Harrison's Architect launches auction-based AI inference marketplace Liquid Inference
- —Providers bid in real time for inference jobs; buyers weigh price, latency and data handling
- —Platform hosts 28 verified providers and more than 1,000 models, including multi-modal ones
- —Provider onboarding takes about 3 hours on average; OpenAI and Anthropic APIs are supported
- —Architect previously raised a $5 million seed round and already offers compute forwards and options
Why it matters: Auction-based pricing could cut AI inference costs and tie the compute market closer to decentralized compute infrastructure.
Source: Crypto Briefing