Trading fees from QTA buy AI inference. Every round, holders get a share as API credit, in proportion to what they held. Spend it through one OpenAI-compatible key, on every model.
Connect your wallet or paste any address to see what it has earned.
Keep at least 100,000 QTA in a wallet you control. Tokens in pools or on exchanges don't count.
We take several snapshots at random times each round. Your share follows the lowest balance you held across them.
Sign a message to prove the wallet is yours. It's free and moves no funds. Then create an API key.
Use the key with any OpenAI-compatible client. Each request bills from your credit at list price.
Each round we claim the trading fees, buy inference with them, and share the credit across every eligible wallet. Here's what each round paid and to whom.
Requests bill from your credit at these list prices. Self-hosted models are labelled, along with the precision they run at.
Change the base URL and the key. Model names, streaming and tool calls work the way they do today. GET /v1/models lists every model id.
# before OPENAI_BASE_URL=https://openrouter.ai/api/v1 OPENAI_API_KEY=sk-or-v1-… # after OPENAI_BASE_URL=https://api.quota.tools/v1 OPENAI_API_KEY=sk-quota-…
Your share of a round's credit pool equals your share of all eligible QTA, using the lowest balance you held across that round's snapshots. The round reports above show exactly what each round paid.
Credit you've already earned stays on your wallet's account. You stop earning in rounds where you don't hold the minimum through every snapshot.
No. You sign a plain text message that proves you own the wallet. It isn't a transaction and can't move funds.
Requests go to the model you name. Upstream responses carry the provider's own model id, and self-hosted models list their precision.
No. Quota keeps billing records only: the model, token counts, cost and timing of each request.