One bill for every model
Video
Image
Text
Gemini 3.8 Flash
Launch promogemini-3.8-flash
Per 1M input and output tokens
Claude Opus 5.5
Launch promoclaude-opus-5.5
Per 1M input and output tokens
Claude Sonnet 5
Launch promoclaude-sonnet-5
Per 1M input and output tokens
Gemini 3.6 Flash
Launch promogemini-3.6-flash
Per 1M input and output tokens
Gemini 3.1 Pro
Launch promogemini-3.1-pro-preview
Per 1M input and output tokens
Grok 4.6
Launch promogrok-4.6
Per 1M input and output tokens
Grok 4.5
Launch promogrok-4.5
Per 1M input and output tokens
Grok 4.3
Launch promogrok-4.3
Per 1M input and output tokens
Kimi K3
Launch promokimi-k3
Per 1M input and output tokens
DeepSeek V4.1 Flash
Launch promodeepseek-v4.1-flash
Per 1M input and output tokens
Official prices checked 2026-10-14. Text requests are capped at 4,096 output tokens.
Why are your prices lower?
We run on cheaper infrastructure and pass the saving on. The price shown is the price you are billed.
How am I billed?
Every model draws from one prepaid API balance. Video and image are charged when a request starts. Text holds its maximum cost, then keeps only the input and output tokens used. Failed generations are refunded automatically.
What is the launch promo?
Create your first API key by Oct 27, 2026 and pay launch prices on every model for 3 months. Your end date is shown in your dashboard.
Is the API balance shared with my Korpi plan?
No. API funds are separate from subscription and MCP credits.
Where do the official prices come from?
Each provider's own pricing page, linked next to every price.