How much does Jev cost?

A typical Jev decision (~1,000 input tokens) costs about $0.0004 — a small fraction of a cent. You're billed simply by input tokens ($0.42 per million, output free) and pay with prepaid credit, not a subscription: no seat fee and no waitlist.

The honest version of the Jev price question: you buy credit up front, and each call to the decide endpoint draws down based on the input tokens it reads — the state plus the questions you send. Output is free. In practice a typical decision is around 1,000 input tokens, so the cost per answer is a small fraction of a cent whether you are classifying a support ticket, scoring a lead, or gating a tool call — larger requests cost proportionally more.

How the Jev price is billed

Credit is sold in packs from $5 to $100, and the larger packs carry bonus credit, so the effective Jev price drops as you scale. There is no monthly lock-in — unused credit stays yours. Grab a jv_live_ key with no waitlist and the same $0.42 per million input tokens applies from your first call. The pricing page has the current pack tiers and bonuses.

• $0.42 per million input tokens — output is free
• Prepaid credit packs from $5 to $100, no subscription and no seat fees
• Bigger packs include bonus credit ($50 and $100 tiers), lowering the effective cost
• No waitlist: the rate is the same from your first jv_live_ call

What one decision costs

Because you pay per input token, the cost of a single decision depends on how much state and how many questions you send. At $0.42 per million input tokens, with output free, here is what one decision runs for common sizes:

• 100 input tokens (a small decision) ≈ about $0.00004
• 500 input tokens ≈ about $0.00021
• 1,000 input tokens (a typical decision) ≈ about $0.00042
• 2,000 input tokens (long text, e.g. moderation) ≈ about $0.00084
• 5,000 input tokens (a large decision) ≈ about $0.0021

Jev cost vs an LLM

Prompting an LLM to make the same typed decision means paying for both input and output tokens on every call, plus the engineering cost of parsing replies and catching invalid labels. With Jev you pay for input tokens only, output is free, and there is none of that overhead — and it lands in roughly 70–500ms, about 40–200x faster than an LLM, so you save on latency as well as on the bill.

Estimating your spend

Size a pack from your input tokens: decisions per month times the tokens each one sends, at $0.42 per million. Since a typical decision is around 1,000 input tokens — about $0.0004 each — even hundreds of thousands of decisions stay modest, and the $50 and $100 packs drop the effective rate further ($0.35/M and $0.25/M). To see the exact tiers and buy credit, open the pricing page, or send a test call to POST https://jevtypesafeai.com/api/v1/decide with a jv_live_ key first.

See also: Pricing & credit packs · How to use the API · Playground

Related: Jev benchmark · Jev models · Jev vs an LLM