The inference pool where GPUs earn and builders save.
Providers register OpenAI-compatible capacity and earn per token. Developers get drop-in, quality-scored inference from $0.30–$1.80 / 1M tokens — no cloud contracts, no lock-in.
- Active providers
- Models online
- $0.30
- From / 1M tokens
Two products, one marketplace loop.
Providers supply capacity. Developers buy reliable, lower-cost inference. Policate manages trust, escrow, routing, and observability between both sides.
Provider loop
Bring capacity, earn per token
- 1Connect capacityOllama, vLLM, or any OpenAI-compatible /v1 endpoint.
- 2Set your termsPick model, price floor, region, and concurrency.
- 3Serve trafficThe pool routes, health-checks, and scores quality.
- 4Get paidEscrow, dispute windows, and Stripe payouts.
Developer loop
Cheaper inference, drop-in API
- 1Add creditsStripe top-up with budget-safe metered usage.
- 2Point your clientDrop-in OpenAI base URL and pool_live key.
- 3Route by qualityTrust scores, retries, fallback, and health.
- 4Spend lessTransparent token pricing, no cloud contracts.
Any OpenAI-compatible engine can join the pool
Capacity-aware routing
Every request matched to the best provider by trust, price, latency, and region — with retries and fallback.
Trust & escrow
Continuous quality spot-checks and fraud detection. Credits held in escrow with dispute windows before payout.
Health & scoring
Live health checks and quality scoring keep slow or degraded providers out of the routing pool automatically.
Metered accounting
Transparent per-token pricing, budget-safe metering, and Stripe payout accounting for both sides of the market.
One base URL. Every OpenAI-compatible client.
Add credits, create a pool_live key, and point your client at the pool. Provider health, retries, fallback, and marketplace accounting stay behind the API.
- No AWS account, no IAM, no cloud contracts
- Capacity-aware routing with automatic fallback
- Budget-safe metered billing via Stripe
export OPENAI_BASE_URL="https://api.providers.policate.marcllort.com/v1"
export OPENAI_API_KEY="pool_live_xxxxx"
curl "$OPENAI_BASE_URL/chat/completions" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen2.5-coder-32b",
"messages": [{"role": "user", "content": "Refactor this function"}]
}'Turn idle GPU hours into revenue.
Register any OpenAI-compatible endpoint — Ollama, vLLM, TGI. Set a price floor, serve pool traffic, and get paid per token after quality checks clear escrow.
1× A100 · qwen2.5-coder-32b · ~62% utilization · $0.74/1M floor
A marketplace, not a contract.
Transparent per-token pricing and quality-scored routing — versus opaque enterprise agreements, committed spend, and single-vendor lock-in.
Questions, answered.
How trust, routing, payouts, and pricing actually work on the pool.
Start with the side of the market you need.
Register capacity as a provider, or use the pool as a developer with a standard OpenAI-compatible API key.