Playground: feel what a judgment model does

Same input: on the left, a generative model's typical output; on the right, Jev's typed judgment. All responses are pre-recorded real API results (jev-1.13.0) with actual token usage. Try editing the input — the response stays fixed, it belongs to the original recorded input.

Primitive · Noul

Is this support message urgent?

Noul evaluates a yes/no question and returns the probability of "yes" — itself a number from 0 to 1.

state (input, editable)

questions (the request)

{
  "urgency": {
    "type": "noul",
    "instructions": "Does this message express urgency?"
  }
}

A generative model would output

I'm really sorry to hear you've been having trouble connecting your Stripe account. I understand how frustrating that must be, especially when it affects your sales. Let me help you look into this. Could you tell me a bit more about...

Jev returnsPre-recorded · not live inference

noul = 0.98

Yes · urgent0.98

{
  "model": "jev-1.13.0",
  "answers": {
    "urgency": {
      "type": "noul",
      "noul": 0.98
    }
  }
}

This call: 301 input tokens (output is free) · 21 output

Noul primitive docs →Confidence routing pattern →

Primitive · Choice

What does this email want?

Choice picks one option from a fixed set, with per-option probabilities and an overall confidence.

state (input, editable)

questions (the request)

{
  "intent": {
    "type": "choice",
    "instructions": "What is the primary intent of this message?",
    "criteria": {
      "bug_report": "Reporting something broken",
      "feature_request": "Asking for new functionality",
      "how_to_question": "Asking how to do something",
      "other": "None of the above"
    }
  }
}

A generative model would output

Thanks for reaching out! It sounds like you are experiencing an error with the export functionality on the reports page when using Safari. This could be related to browser compatibility. Have you tried...

Jev returnsPre-recorded · not live inference

choice = bug_report · confidence 1.0

Bug report1.00
How-to question0.02
Other0.01

{
  "model": "jev-1.13.0",
  "answers": {
    "intent": {
      "type": "choice",
      "choice": "bug_report",
      "confidence": 1,
      "probabilities": {
        "bug_report": 1,
        "feature_request": 0,
        "how_to_question": 0,
        "other": 0
      }
    }
  }
}

This call: 389 input tokens (output is free) · 50 output

Choice primitive docs →Intent routing pattern →

Primitive · Score

How severe is this issue?

Score rates on ordered criteria; the value can land between levels, with per-level probabilities and confidence.

state (input, editable)

questions (the request)

{
  "severity": {
    "type": "score",
    "instructions": "How severe is the issue?",
    "criteria": [
      "Minor",
      "Moderate",
      "Critical"
    ]
  }
}

A generative model would output

I'm so sorry to hear that your payouts have been failing for three days. That sounds really stressful. Payout failures can have several causes, such as... (510 more words)

Jev returnsPre-recorded · not live inference

score = 1.87 (between Moderate and Critical) · confidence 0.81

Minor0.00
Moderate0.13
Critical0.87

{
  "model": "jev-1.13.0",
  "answers": {
    "severity": {
      "type": "score",
      "score": 1.87,
      "confidence": 0.81,
      "legend": {
        "0": "Minor",
        "1": "Moderate",
        "2": "Critical"
      },
      "probabilities": {
        "0": 0,
        "1": 0.13,
        "2": 0.87
      }
    }
  }
}

This call: 321 input tokens (output is free) · 36 output

Score primitive docs →Composite scoring pattern →

Pipeline pattern

Composite scoring: how fast should we respond?

Combine two independent scores — emotional intensity and business impact — into a weighted response priority. All typed, all explainable.

state (input, editable)

questions (the request)

{
  "emotional_intensity": {
    "type": "score",
    "instructions": "How emotionally intense is this message?",
    "criteria": [
      "Calm and factual",
      "Frustrated but professional",
      "Angry or escalating"
    ]
  },
  "business_impact": {
    "type": "score",
    "instructions": "How severe is the business impact described?",
    "criteria": [
      "No material impact",
      "Degraded operations",
      "Revenue-blocking outage"
    ]
  }
}

A generative model would output

Dear valued customer, thank you for bringing this to our attention. We sincerely apologize for the inconvenience caused by the API errors. Our engineering team is investigating the issue with the highest priority...

Jev returnsPre-recorded · not live inference

priority = 0.6×2.0 + 0.4×1.82 = 1.93 → respond immediately

Emotion · escalating0.82
Impact · revenue-blocking1.00

{
  "model": "jev-1.13.0",
  "answers": {
    "emotional_intensity": {
      "type": "score",
      "score": 1.82,
      "confidence": 0.74,
      "legend": {
        "0": "Calm and factual",
        "1": "Frustrated but professional",
        "2": "Angry or escalating"
      },
      "probabilities": {
        "0": 0,
        "1": 0.18,
        "2": 0.82
      }
    },
    "business_impact": {
      "type": "score",
      "score": 2,
      "confidence": 1,
      "legend": {
        "0": "No material impact",
        "1": "Degraded operations",
        "2": "Revenue-blocking outage"
      },
      "probabilities": {
        "0": 0,
        "1": 0,
        "2": 1
      }
    }
  }
}

This call: 397 input tokens (output is free) · 37 output

Composite scoring pattern →LLM guardrails case →

Pipeline pattern

Fan-out: every judgment in one call

No classify-then-follow-up. Send every question you might need at once — answers are independent, your code decides which to use. Output tokens are free.

state (input, editable)

questions (the request)

{
  "request_type": {
    "type": "choice",
    "instructions": "What kind of request is this?",
    "criteria": {
      "sales_question": "Asking about plans, pricing, compliance, or contracts",
      "billing_issue": "Payment, invoice, or refund matters",
      "technical_support": "Something is broken or needs troubleshooting"
    }
  },
  "is_existing_customer": {
    "type": "noul",
    "instructions": "Does the writer appear to be an existing customer?"
  },
  "tone_politeness": {
    "type": "score",
    "instructions": "How courteous is the tone overall?",
    "criteria": [
      "Blunt or rude",
      "Neutral",
      "Warm and appreciative"
    ]
  }
}

A generative model would output

Thank you for reaching out! To answer your questions: regarding SOC 2 report access, our enterprise plan does include... (the answer continues for 300+ words, mixing both topics with no typed structure to route on)

Jev returnsPre-recorded · not live inference

choice=sales_question (conf 0.25) · noul=0.86 · score=1.74 — one call, three independent judgments

Type · sales 0.500.50
Existing customer · yes 0.860.86
Tone · warm 0.740.74

{
  "model": "jev-1.13.0",
  "answers": {
    "request_type": {
      "type": "choice",
      "choice": "sales_question",
      "confidence": 0.25,
      "probabilities": {
        "sales_question": 0.5,
        "billing_issue": 0.5,
        "technical_support": 0
      }
    },
    "is_existing_customer": {
      "type": "noul",
      "noul": 0.86
    },
    "tone_politeness": {
      "type": "score",
      "score": 1.74,
      "confidence": 0.61,
      "legend": {
        "0": "Blunt or rude",
        "1": "Neutral",
        "2": "Warm and appreciative"
      },
      "probabilities": {
        "0": 0,
        "1": 0.26,
        "2": 0.74
      }
    }
  }
}

This call: 459 input tokens (output is free) · 78 output

Fan-out pattern →Parallel questions case →