1 live run · no signup
Jev PlaygroundRun the Jev model online, right in your browser
Paste a message, pick a question and get a typed answer from the live typesafe/jev-1.13 model, with the probability behind it. Your first run needs no account. Verify an email for 5 free runs every UTC day.
Comments
moderation + spam
Live demo
POST /api/jev/runs
Answer
illustrative answer
toxic
Support inbox
intent + routing
Side panels show illustrative answers. The middle one runs the live model.
How to use the Jev playground
01
Pick a template or paste your own text
Choose the comment moderation or support routing template, then replace the sample with a short, non-sensitive message of your own. This page accepts up to 600 characters.
02
Run it on the live model
Press Decide. JevModel sends your text and the template question to typesafe/jev-1.13 through OpenRouter and shows the selected label with its probability. Side panels show illustrative answers; the middle panel is live.
03
Write your own questions in the Console
Sign up and verify your email to open the full Playground. There you can send text or JSON state with up to 8 Choice, Score or Noul questions per request, and you get 5 free successful runs each UTC day.
04
Move to the API, Batch or Decision Flows
When a question works, call the same request from your server with the Jev API, run a CSV through Batch, or chain questions into a Decision Flow.
What you can test in the playground
Choice
Selects one label from the options you name, such as billing, technical or sales. Returns the label and a probability for each option.
Score
Rates the input on an ordered rubric you write, such as lead priority from 1 to 5. The score can be fractional.
Noul
Returns the probability that a statement is true, such as "a person should review this case now".
Jev playground options
TypeSafe AI, which develops Jev, runs its own playground in the TypeSafe console. JevModel is an independent playground that calls the same model through OpenRouter. Pick the one that fits how you work.
| TypeSafe console playground | JevModel playground | |
|---|---|---|
| Operator | TypeSafe AI (first party) | Independent; not affiliated with TypeSafe AI |
| Model | Jev, served by TypeSafe | typesafe/jev-1.13 via OpenRouter |
| Account | TypeSafe console login | None for 1 run a day; email verification for 5 free runs a day |
| After the playground | TypeSafe API and official SDKs | JevModel API, browser Batch and visual Decision Flows |
| Billing | TypeSafe billing | Free daily runs, then token packs that never expire |
Jev playground FAQ
Is the Jev playground free?
Yes. You can run the demo on this page once per UTC day without an account. After you verify an email, you get 5 free successful runs every UTC day, subject to site capacity. Additional runs use a token balance.
Is this the official TypeSafe playground?
No. JevModel is an independent service and is not affiliated with TypeSafe AI. Live answers come from the real typesafe/jev-1.13 model through OpenRouter; they are not simulated.
Are my playground inputs stored?
No-signup runs are not saved to any history. Signed-in runs are kept in your Console history for up to 30 days (latest 100) and you can delete them. Inputs are sent to the model provider to produce an answer, so do not paste secrets or sensitive personal data.
What are the input limits?
This page accepts up to 600 characters. The Console Playground and API accept text, a JSON object or a JSON array of up to 8,000 characters, with 1 to 8 questions per request.
Can the Jev playground write text or chat?
No. Jev answers bounded questions with labels, scores and probabilities. Use a text-generation model when you need written replies.