Models
Model API pricing
Model API pricing: DeepSeek V4.1 Flash direct $0.11 in / $0.23 out per million tokens; Sansi $0.30 in / $0.60 out per million user-visible tokens; the prompt guard is billed by estimated tokens.
| Model | Billing unit | Price (USD) |
|---|---|---|
| deepseek-flash (direct) | tokens | 0.11 in / 0.23 out per million tokens |
| deepseek-flash-three-pass (Sansi) | User-visible tokens: the request plus the final reply. The internal three rounds are not billed separately | 0.30 in / 0.60 out per million tokens |
| xjp-guard-1 (prompt guard) | Estimated tokens: one unit per 400 tokens, plus 1 base unit | 0.003 per unit |
Sansi is deliberately not billed for its internal three rounds: you pay only for the tokens you can see, and orchestration overhead is covered by the rate. The rate may change as the benchmark changes, and there is no separate tuning fee.
Prompt guard examples
| Input | est_tokens | units | Price (USD) |
|---|---|---|---|
| One English sentence (48 characters) | 12 | 2 | 0.006 |
| English, 2,000 characters | 500 | 3 | 0.009 |
| Chinese, 2,000 characters | 2,000 | 6 | 0.018 |
| English, 8,000 characters | 2,000 | 6 | 0.018 |
| Chinese, 8,000 characters | 8,000 | 21 | 0.063 |
Request an API key
Self-service signup, credits, and online payment are still being connected. For now, request a key by email and tell us what you will use it for and your expected volume.
Listed prices include tax and invoicing. There are no add-on fees.