OpenJev · Laya 1.0
Give Laya a short message and a question. It can pick an answer, say yes or no, or give a score.
“I was charged twice.”
“Which team should help?”
For example: “Billing.”
Your idea. Your questions.
Write the text you want Laya to read. Add a question and tell it what kind of answer you need. Then use your LLMRegistry API key to get a real result.
Prepare your text and questions, then run your query.
Laya will give one result for each question. You can also see the exact information you send and the JSON it returns below.
Laya can make mistakes, even when its confidence is high. Check the result before taking an important action.
The payload updates as you build your use case. Your API key is never included here.
Preparing your payload…Run your query to see the result.Replace YOUR_LLMREGISTRY_API_KEY with your key, then run this command in a terminal.
See it in action
Choose an example. Follow the message, question, and result.
Example 01
A help desk can send each message to the right team.
You can edit this text, then use your API key to run it.
Which team should handle this message?
Loading…
Laya can make mistakes, even when its confidence is high. Check the result before taking an important action.
Loading the recorded examples…
Under the hood
An API lets one app ask another app for help. The payload is the information you send. JSON is a way to label that information so the app can read it.
Loading payload…Loading result…Replace YOUR_LLMREGISTRY_API_KEY with your key. Run this command in a terminal to send the payload.
choice is the answer Laya picked. probabilities shows how much it favored each answer. A value of 0.8 means 80% of the answer weight.
noul is Laya’s estimate for “yes.” Near 1 means it favors yes. Near 0 means it favors no. Near 0.5 means it is unsure.
The levels start at 0. With unhappy, neutral, and happy, the scores are 0, 1, and 2. The result is a weighted average, so Laya can give a score such as 1.7.
You can ask up to 4 questions in a request, with up to 10 choices per question. Laya reads text and returns decisions. It reads up to 1,024 tokens per question; a token is a small piece of text. Longer text may be cut short. Each question counts the message again in usage.input_tokens. These decisions use 0 output tokens.
Saved examples are labeled. A live call uses your API key and the message you entered. If the service is busy, wait a little and try again.