Overview
Signed in as …
Get started
0 of 2 done
-
Create an API key (not done)
A key is shown only once, so store it somewhere safe.
-
Open PlaygroundMake your first successful request (not done)
Try it in the Playground, or call the API from your code.
Show code
Keep your key in an environment variable:
export JEV_API_KEY="jev_…"curl https://api.jevstyle.com/v1/systemone \ -H "Authorization: Bearer $JEV_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "state": "I was billed twice. Please refund one charge.", "questions": { "refund": {"type": "noul", "instructions": "The customer asks for money back."} } }'import os import requests response = requests.post( "https://api.jevstyle.com/v1/systemone", headers={"Authorization": f"Bearer {os.environ['JEV_API_KEY']}"}, json={ "state": "I was billed twice. Please refund one charge.", "questions": { "refund": {"type": "noul", "instructions": "The customer asks for money back."}, }, }, timeout=70, ) response.raise_for_status() print(response.json()["answers"]["refund"]["noul"]) # probability of "yes"
jev-style-decision-v3Service health
Checking…- APIChecking…
- InferenceChecking…
Resources
- DocumentationRequest format, question types and examples
- EndpointsEvery route, with request and response fields
- ErrorsStatus codes and how to handle them
- Limits & modelsRate limits, daily quota and model details
Playground
Build decision questions and run them against POST /v1/systemone. Successful runs count toward your daily quota.
Kept in this tab's memory only — never saved. Sent only to the JevStyle API.
Questions
The full systemone request, including state.
API keys
A key is shown only once, when you create it, so store it somewhere safe. Do not share your keys or put them in browser or other client-side code. All keys in your account share the same limits.
| Name | Secret key | Created | Last used | Requests today | Actions |
|---|---|---|---|---|---|
| Loading… | |||||
Usage
Successful requests across your API keys. Days follow UTC.
Successful requests
Last 30 days · UTC
Show as table
| Day (UTC) | Requests | By key |
|---|
- Remaining
- —
Only successful requests count. Failed, rejected or timed-out requests are not charged. Requests made with keys you have since revoked still count toward today's total.
Limits & models
Free tier, best effort and without an SLA. Limits may change.
Rate limits
Limits apply to your whole account and are shared by all of its keys.
| Model | Concurrent | Requests per minute | Requests per day | Input per request |
|---|---|---|---|---|
jev-style-decision-v3 |
1 | 20 RPM | 10,000 RPD | 8,000 tokens |
- The daily quota resets at 00:00 UTC. Only successful requests count; failed, rejected and timed-out requests are free.
- Over a limit, requests get
429(withRetry-Afterfor the per-minute limit). Input over the token cap gets413. - Up to 3 active API keys per account. Extra keys are for rotation, not extra capacity.
Models
jev-style-decision-v3
Free tier
Jev-Style Decision v3 reads a state and answers typed decision questions in a single pass, returning calibrated probabilities instead of generated text.
- Context window
- 25,600 tokens
- Free-tier input cap
- 8,000 tokens per request
- Question types
noulyes / nochoicepick one optionscoreordered scale- Output
- Probabilities per question, no generated text
- Endpoints
POST /v1/systemoneGET /v1/models