Skip to content
Laya Model Online

Try Laya AI Online for Text Classification

Turn support messages and customer feedback into a label you defined, with a probability for each category. Start from a task, edit the rules, and read the result before you route anything.

No installation required. Start with a sample or use your own text.

Try a support ticketExplore templates

Live sends the text to this site's server, then to the official Laya Hugging Face Space. No account is required for one message.

Rough budget: about 4 of 320 tokens for the text, and about 87 of 180 for the questions. The English checkpoint context is 512 tokens, with about 192 reserved for the question. This estimate is characters divided by 4, not the real tokenizer.

Categories

The key is the label that comes back. The description tells the model what belongs there.

No result yet

Load an example or paste a message, then choose Run classification. Loading an example does not send it.

Saving rules to an account is not available. There is no signup. Copy the rules if you want to keep them.

What can you do with Laya here?

Classify support ticketsAssign one incoming request to categories you name, such as billing, technical support, sales, or other.Organize customer feedbackLabel one note as a bug, feature request, usability issue, pricing comment, praise, or other.Test decision rulesOpen the playground, edit the JSON, and inspect the raw response. Same inference path as the tools.

How to use Laya online

  1. Add your text. Paste one message, review, or ticket, or load an example. Loading does not run it.
  2. Choose the decision. Keep the template categories or rewrite the names and descriptions.
  3. Run and review. Read the label and the probabilities, then change the rules and test another example.

The longer version is the how to use Laya guide.

What Laya is, in short

Laya is a decision model. It scores questions you define. It does not draft the customer reply. A chat model writes text. Laya returns a choice, an ordered score, or a yes/no probability. Predictions still need evaluation on your own examples before you automate a queue. Read what Laya is for the checkpoints and the benchmark limits.

Three recorded ticket calls

These are single calls, not testimonials and not a test-set score. One call each on 2026-09-29 to the official Space convaiinnovations/laya-demo, playground endpoint, state {"message": "..."} and the question shown on that tool. The model field came back as laya. latency_ms is the Space response field, not this website's end-to-end time and not the upstream GPU figure. A later call can differ. These six calls are not an accuracy score.

A clear refund request

I paid twice for order 4821. Could you refund the extra payment?

Human reference. Billing. A reviewer would send this to the team that handles payments and refunds.

This call returned. billing. Model field laya. Space-reported latency 112.9 ms. Input tokens 84. Confidence 0.6371. Date 2026-09-29. Space convaiinnovations/laya-demo.

  • billing: 0.8777
  • other: 0.0498
  • sales: 0.0434
  • technical: 0.0291

The label matches the human reference on this one message. That is not a measured accuracy rate for the template.

A crash and a duplicate charge

The app crashes when I open invoices, and I was also charged twice. Can someone fix the crash and refund the extra charge?

Human reference. Both technical and billing. The question asks for one team. A reviewer could defend billing because of the refund, while the crash still needs technical help.

This call returned. billing. Model field laya. Space-reported latency 123.2 ms. Input tokens 96. Confidence 0.2999. Date 2026-09-29. Space convaiinnovations/laya-demo.

  • billing: 0.5944
  • technical: 0.3036
  • other: 0.0656
  • sales: 0.0364

A single choice cannot hold two requests. Technical still received 0.3036. The low confidence figure is distribution concentration, not a calibrated chance that billing is correct.

A negation about a refund

Please do not refund order 4821. The second charge was intentional and covers next month.

Human reference. The payment topic still belongs with billing, but the customer is refusing a refund. Billing here must not be read as 'refund requested'.

This call returned. billing. Model field laya. Space-reported latency 103.5 ms. Input tokens 87. Confidence 0.4063. Date 2026-09-29. Space convaiinnovations/laya-demo.

  • billing: 0.7556
  • other: 0.0923
  • sales: 0.0826
  • technical: 0.0696

This call does not show that the model understood 'do not refund'. The billing description mentions refunds, so the word refund can pull the label. Add a separate yes/no question if you need the intent, and review negations before acting.

Who runs this, and where the text goes

Laya Model Online is an independent tool powered by Laya. We are not affiliated with or endorsed by the Laya project or Convai Innovations. Live classification sends your input to our hosted server and then to the official Hugging Face Space. It does not stay on your device. The offline demo does. Details are on the privacy page.

What you can do without an account

Classify one message, edit the categories, copy the decision rules, and read the raw JSON in developer mode. Saved classifiers, history, CSV batch, and a paid API are not available, so this site has no price table. The upstream package is Apache 2.0. That license is not a bill for this website, and it is not a promise that hosted runs stay free.

Questions

What is Laya?
Laya is a System-1 decision model from Convai Innovations. You give it text and questions you define. It returns a label, a score, or a yes/no probability. It does not write a chat reply.
Can I use Laya online without installing it?
Yes. Live mode on this site sends one message to the official Convai Hugging Face Space convaiinnovations/laya-demo. You can also install the Apache-2.0 package yourself with pip install laya. This website does not download the weights.
Is this the official Laya website?
No. Laya Model Online is an independent unofficial tool. It is not affiliated with or endorsed by Convai Innovations. It is also not LayaAir, the Layabox game engine.
Is it free to use?
One message at a time can be classified without an account. Saving rules to an account, CSV batch, and an API product are not offered. There is no price list because those features are not for sale here. The Hugging Face Space has its own queue and limits.
What data can I submit?
Short text, or JSON in developer mode. This site blocks a live run when a rough token estimate exceeds the English checkpoint budget, about 320 tokens for the text. Do not submit secrets you cannot share with this host and with Hugging Face.
Which languages are supported here?
The site interface is English. The public Space router can send non-English text to the multilingual checkpoint, but this site has not validated those routes. “100+ languages” describes the multilingual checkpoint, not a promise about this website.
What does the probability mean?
It is the model’s predicted probability for a label you defined. It is not an accuracy rate and not a fact. Confidence is how peaked that distribution is. Do not treat a high number as safe to automate.
Can I classify multiple records?
No. Each run is one message. CSV batch is not available.
Is my input stored?
This application does not write the message to its own database or logs. Live mode still sends the text to our server and then to the Hugging Face Space, so it leaves your device. Offline demo stays in the browser. See the privacy page.
Can I automate actions with the result?
Not from this site. Review uncertain or consequential decisions before you act. Laya returns a prediction. It does not grant permission to refund, route, or reply.