Updated

Published

Venice lists Jev on its API: same key, no copy of the state

Venice put Jev on a beta Decisions API the night of September 18. Calls go to POST /api/v1/decisions with model jev-latest, billed at TypeSafe's $0.042 per million input tokens with free output. Venice says it keeps no copy of the state or the answers, and TypeSafe quoted the launch.

Venice posted on September 18 that Jev is on its API, in beta. TypeSafe quoted the announcement.

The landing page at venice.ai/lp/jev is the spec. Calls go to POST https://api.venice.ai/api/v1/decisions with model jev-latest. The three question types are the TypeSafe ones: noul, choice, and score. Price matches TypeSafe’s public rate, $0.042 per million input tokens, output free. Limits match TypeSafe’s models page: 64,000 tokens for the whole request, 32,000 for state plus the longest question.

The same Venice key that reaches chat, image, video, and audio reaches Jev. Venice says it keeps no copy of the state or the answers, and that it does not train on that traffic. Auth is a Bearer key, or an x402 wallet with prepaid USDC credits. The same handler is also served at /systemone, so a TypeSafe SDK can change the base URL and keep the rest of the call.

Jev is API only. It has no surface in the Venice web app. The page says availability and schemas can change without notice, and that GET /models?type=decision is the list to trust.

Venice’s thread suggests a confidence split for choice and score: act at 0.90 or above, ask for confirmation at 0.60, send the rest to a person. Those cutoffs are examples on the landing page, not a TypeSafe setting. Noul still returns a single probability from 0 to 1, with no separate confidence field.

New Venice accounts, the page says, include a free daily allowance and 500 welcome credits, with no credit card required. We did not create an account or send a request.

This site's reading

Editorial notes evaluating claims against primary sources, contextualizing findings alongside related implementations, and defining technical terms.

Verify

Venice's September 18 thread and the public landing page agree: Jev is in beta on POST https://api.venice.ai/api/v1/decisions with model jev-latest, $0.042 per million input tokens, free output, 64K tokens for the whole request and 32K for state plus the longest question. The page says the same Venice key reaches chat, image, video, audio, and Jev; that Venice keeps no copy of the state or the answers and does not train on them; that auth is a Bearer key or an x402 wallet with prepaid USDC; and that the same handler is also served at /systemone so a TypeSafe SDK can swap the base URL. Schemas are not frozen. Jev is API only, with no surface in the Venice web app. We did not call the endpoint.

Compare

OpenRouter, Cloudflare, Netlify, and Vercel already list Jev. Venice is the first of those doors that leads with no retention and with a wallet path. OpenRouter's Decisions API is still labeled alpha (`/api/alpha/decisions`). TypeSafe's own docs use POST /v1/systemone and the same 64K / 32K split. The 0.90 / 0.60 confidence thresholds on the Venice page are examples, not TypeSafe policy; Noul still has no separate confidence field, which matches TypeSafe's primitives.

Terms

Decisions API
Venice's endpoint POST /api/v1/decisions, which accepts state and typed questions rather than a chat transcript. OpenRouter uses the same name for POST /api/alpha/decisions.
x402 wallet
A prepaid USDC path Venice lists next to a Bearer API key for authenticating decision calls.
No state retention
Venice's claim that it keeps no copy of the state you send or the answers it returns, and that it does not train on that traffic.

Sources

  1. Venice, Jev on the Venice API in beta
  2. Venice thread on pricing, 64K, and the endpoint
  3. TypeSafe quoting the Venice listing
  4. Venice landing page for Jev