Billed per million tokens.
Three model lanes, one API contract.
Pay for what you use.
| Model ID | Input Window | Output Window | Input / 1M Tokens | Output / 1M Tokens | Status |
|---|---|---|---|---|---|
| lucena-coder-latest | 80,000 | 8,000 | $4.00 | $8.00 | Preview |
| lucena-coder-roadrunner | 24,000 | 4,000 | $1.00 | $2.00 | Preview |
| lucena-pulse-latest Beta | 64,000 | 8,000 | $0.80 | $2.00 | Private Beta |
Tokens are fragments of words, roughly 4 characters each. A typical English sentence is about 15 to 20 tokens. Code is denser: a 100-line function might be 400 to 600 tokens.
Yes. All tokens processed by the model, including any system prompt, count toward input tokens. Keep system prompts concise to reduce costs.
Yes. Preview limits are 60 requests per minute and 200,000 tokens per minute. Contact us for higher limits or dedicated capacity.
No. All sessions are ephemeral. We do not log, store, or train on any request or response content. See our privacy policy.
For volume commitments, dedicated instances, or custom SLAs, use the Enterprise link in the footer or email enterprise@lucena.one.