Architect launches Liquid Inference, a real-time auction for LLM inference

Architect Financial Technologies has launched Liquid Inference, an LLM router that runs a live auction for every request, with providers bidding to serve each prompt and buyers paying the lowest qualifying offer. Max price is locked before generation, and buyers can set cost caps, latency limits and region or data-retention rules.

• Providers compete per request; the lowest qualifying offer wins
• Max price is locked before generation, billing covers metered usage only
• Drop-in compatible with OpenAI and Anthropic APIs, including Claude Code and Cursor
• First 500 users get $20 free inference; referrals earn 20% of fees