{"service":"hivecompute","version":"2.0.0","description":"Inference router for the Hive agent economy. Context compression + model routing on every call.","endpoints":{"inference":"POST /v1/compute/chat/completions","models":"GET  /v1/compute/models","estimate":"POST /v1/compute/estimate","usage":"GET  /v1/compute/usage/:did","compression":"GET  /v1/compute/compression/stats","embeddings":"POST /v1/compute/embeddings","mcp":"POST /mcp","benchmark":"GET  /v1/compute/benchmark","benchmark_semantic":"GET /v1/compute/benchmark/semantic","benchmark_temporal":"GET /v1/compute/benchmark/temporal","benchmark_modality":"GET /v1/compute/benchmark/modality"},"benchmark_note":"Three public triangulated inference compression benchmark nodes. No auth required. DID-signed. Live production data. Cite freely. https://hivecompute-g2g7.onrender.com/v1/compute/benchmark","pricing":"USDC per token on Base L2. Auto-routing enabled by default. Compression runs on every call.","network":"Hive Civilization — hiveagentiq.com","agent_onboarding":"External agents: send X-Agent-DID header on read endpoints for 100 free reads/day per DID. Anonymous reads are $0.001 USDC each."}