New model · Same request as Jev
Mercury Decide on Jev AI
Mercury Decide is Inception’s decision model. Give it a state and typed questions and it returns a choice, a score or a yes/no probability, with the confidence read from the model itself. It uses Jev’s request format, so you can switch any playground run, saved judge or API call by changing one field.
Checking connection…
What Mercury Decide is
Inception builds the Mercury family of diffusion language models, which refine many positions in parallel instead of writing one token at a time. Mercury Decide applies that to decisions: the answer is a probability over a short set of labels, not generated text. Inception quotes up to 14 decisions per second and aims it at moderation, routing, fraud triage, agent tool selection, grading and escalation gates.
On Jev AI it runs alongside Jev, Laya and Jev-Omni. Choose it in the playground model menu, or set model to mercury-decide in the API.
How closely it matches Jev
We replayed every recorded Jev request on this site — 41 requests from the homepage demos and use-case templates — on Mercury Decide and compared the answers with the recorded jev-1.13.0 results. This measures agreement, not accuracy: the examples have no ground-truth labels.
Measured 2026-10-01 on snapshot inception/mercury-decide-20260930, one request at a time; median wall-clock 0.67s from one client, including the network hop. Full comparison and method →
Use Mercury Decide with your Jev API key
Create a key on the Jev API page and store it as JEV_AI_API_KEY on your server. Use the Jev endpoint with mercury-decide, or the Decisions-compatible endpoint with inception/mercury-decide. Check authenticated GET /api/v1/models for availability first.
curl --fail-with-body https://jev-ai.pro/api/v1/systemone \
-H "Authorization: Bearer $JEV_AI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "mercury-decide",
"state": "Customer: I was charged twice for order 4821 and want my money back today.",
"questions": {
"team": {
"type": "choice",
"instructions": "Which team should handle this message?",
"criteria": {
"billing": "Payments, charges and refunds",
"shipping": "Delivery problems",
"technical": null
}
},
"urgent": {
"type": "noul",
"instructions": "Does this need a reply within the hour?"
}
}
}'Read answers.team.choice for the routed team and answers.urgent.noul for the probability of yes. Saved judges work too: send judgeId with model set to Mercury Decide.
Limits and billing
- Up to 32,768 tokens per request, half of Jev’s 64k. Longer input is rejected without a charge; it is never silently truncated.
- Billing is the same as Jev: one credit per playground run, and API calls use your paid input tokens first, with one credit covering the request when tokens are insufficient. Failed requests are not charged.
- Mercury Decide is in early access from Inception. If it is unavailable the request returns an error before any charge; it is never quietly answered by a different model.