Give Laya a short text and a typed question. Get a yes/no probability, a category or a score. Choose English or Multilingual in the playground, then use the same model through the JevTypeSafe API.
Ultra-fast classification model optimized for short English messages, comment moderation, and real-time LLM guardrails.
Direct cross-lingual intent routing and sentiment scoring for global customer support tickets and social media comments without translation.
Select your model and send the same JSON structure to the decisions endpoint.
curl https://jevtypesafe.net/api/v1/decisions \
-H "Authorization: Bearer $JEVTYPESAFE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "laya-english",
"state": "Payment failed since yesterday. Please help.",
"questions": {
"billing": {
"type": "noul",
"instructions": "Is this message about billing or payments?"
}
}
}'Running the weights on your own machine? See Laya local.
| Feature | Jev (Flagship) | Laya English | Laya Multilingual |
|---|---|---|---|
| Context window | 8,000 tokens | 512 tokens | 1,024 tokens |
| Ideal input | Long documents, tickets, citation pairs, nested JSON | Short messages, reviews, moderation comments | Global multilingual support tickets, foreign feedback |
| Average latency | ~0.45s | ~0.28s | ~0.34s |
| Output formats | Choice · Yes/No · Score | Choice · Yes/No · Score | Choice · Yes/No · Score |
| Token billing | Standard Token balance | Standard Token balance | Standard Token balance |