Thirdeye AI pay-per-call inference

Pay for the call, not the seat.

Thirdeye routes requests to several model providers behind one OpenAI-compatible API and bills each call at cost plus margin. No subscription, no seat licence, no minimum.

One API, several providers

Point an existing OpenAI client at our base URL. Model names resolve to a live provider pool, with a fallback chain behind each one.

Per-call pricing

Every JSON response reports its billed amount in x-thirdeye-cost-usd, measured. A stream reports the reserved estimate up front, settles to the measured cost as it ends, and says which figure it is sending in x-thirdeye-cost-basis.

Prepaid credit

Fund the account, or fund a specific API key. A key can carry its own balance and its own monthly ceiling.

Bring your own key

Supply a provider key and it is stored encrypted, used only for your calls, and never shown again.

Agents with budgets

Registered agents run with a step cap, a spend cap and an allowlist for outbound paid tool calls.

Streaming and tools

Server-sent events and OpenAI-shaped tool calling, both billed the same way as a normal call.

Use it

curl https://thirdeye-ai-alpha.vercel.app/api/v1/chat/completions \
  -H "authorization: Bearer $THIRDEYE_API_KEY" \
  -H "content-type: application/json" \
  -d '{"model":"gpt-oss-120b","messages":[{"role":"user","content":"hi"}]}'

Sign in with a wallet to create a key, then fund it with USDC on Base. Full details in the API guide.