Upload your examples.
A labeled TXT file is all you need. Add one example and its 0 or 1 label on each line.
Meet NULIQ Chat: ask, write and explore in your language.
Stream responses from our self-hosted assistant.
Qwen-powered chat, alongside NULIQ decision tools.
NULIQ Core · TXT training and testing · Current pricing in your console · Free output
“I approve this action.”
Make text classification a building block. Let your examples define what “yes” and “no” mean.
A labeled TXT file is all you need. Add one example and its 0 or 1 label on each line.
Train from your console. Measure errors with a separate TXT file. Activate the new version when you are ready.
Send text from your server and get one bit back. API access, Playground, and model management share one account.
One API. One bit. Use the current NULIQ Core or train and activate a separate private model for your task. Arduino and ESP32 devices connect through the API. Answers can be wrong; test your task on separate examples.
For product teams, developers and operators who need a decision their software can use.
Use our neural model to estimate whether an email or phone number has a plausible format. Review uncertain results before accepting a contact.
AI FORMAT → 0 / 1 + SCOREExperiment with billing, technical support and sales routing. Use descriptive options and evaluate on real tickets before automating.
TEXT → CHOICE + SCORESScore requests against your ordered urgency labels. Send uncertain or consequential decisions to a reviewer.
TEXT → SCORE 0–1Use structured responses directly in forms, dashboards and internal tools. Keep your integration small with one REST API.
ONE REQUEST → TYPED JSONConnect Arduino, ESP32 and gateways through HTTPS. Your device sends text; NULIQ runs the model on our server.
DEVICE → GATEWAY → APIUpload labeled TXT examples or verified feedback to train a separate private binary model. Test it before activation.
YOUR EXAMPLES → PRIVATE MODELFast Core is experimental. AI contact scores can be wrong and do not verify existence or ownership.
For small teams managing forms and customer workflows: uploads, evaluation, version selection, and an API in one console.
For example, label a message to be passed to sales as 1 and other messages as 0. Correct labels and representative test data depend on your process.
Train on your examples.Create a separate version for each customer and decision task.
Test on separate data.See false positives and false negatives separately. Tests that overlap with training data are rejected.
Measure, then activate.Your API address stays the same. If a new version does not work for you, select a previous one.
Know the current limits. A yes/no prediction is not a guarantee. Email or phone format checks do not confirm that an address exists, a number is reachable, or an identity is genuine. Guard EN remains a separate, narrow English SMS spam example.
Connect from your server with the REST API. A compatibility endpoint is also available for supported clients using the Ollama API format.
Send JSON.
Get 0 or 1 as plain text.
Update your model.
Keep your integration.
Measure the input.
Pay nothing for output.
# One input. One decision.
curl $NULIQ_URL/v1/decide \
-H "Authorization: Bearer $NULIQ_KEY" \
-H "Content-Type: application/json" \
-d '{"input":"I approve this action."}'Example output · The response body is a single character.
Loading current pricing from the server. Input is measured in UTF-8 bytes; output is free.
1 MB = 1,000,000 bytes · USD
100,000 input bytes—
Some characters use more than one byte. The console shows the exact input size and estimated cost before you send a query.
API and Playground use the same pricing. Failed or invalid queries are not charged. Training and quality tests do not debit your query balance; usage limits apply.
Before paying, review the terms of use, refund and cancellation policy and service delivery information.
Start with NULIQ Core.
Train a private version for your own classification task.
NULIQ provides self-hosted AI chat and dedicated decision APIs. No physical products are shipped. Loading service status.
Current business and support details are on the contact page.
Contact and business details
CONSOLE / My accountChat with NULIQ. Responses stream as they are generated.
Use our own neural model for decisions and contact-format estimates. Enable private automatic learning below to learn from your verified corrections. This panel uses Core and your enabled learned adapter. Read the docs →
Save a verified answer after a decision. When enabled, automatic learning trains a private adapter in the background and activates it only when it improves held-out results. It never treats its own predictions as correct labels.
For each decision or contact head: at least 40 training and 20 held-out examples. Later cycles need 20 new or corrected labels. Promotion requires at least 90% on your held-out set, two additional correct answers, and extra false-positive checks for contacts. Small test sets do not guarantee future accuracy.
Turning this off returns Core to its shared weights. Reset removes automatic-learning examples and adapters. Saved TXT-training feedback is managed separately above. No examples are sent to an external AI provider.
Upload a TXT file to train a separate private model. Follow its progress, test it, then activate it.
No training jobs yet.
TXT and verified feedback train a separate private model on the NULIQ server; they do not retrain hosted NULIQ Core. Your current model keeps running while the new one trains. Test the completed version before activating it. A training score is not an independent test result.
Evaluate a local model on labeled examples that were not used in training. This does not change your active model.
NULIQ Core is tested one question at a time in Playground or through the API. Bulk TXT testing is available for your private models and Guard EN.
This result describes only the test set you uploaded. Percentages are not the model’s overall accuracy or confidence score. You decide whether to activate it.
Create an API key. Use it securely on your server.
Copy it somewhere secure. It cannot be displayed again after you leave this view.
Do not put your API key in public website JavaScript, a mobile app bundle, or a source repository. Send requests through your own server. A key can access all models in your account; model-scoped keys are not available.
Successful response: text/plain, just 0 or 1. model: "active" selects your active version. Use the model ID copied from Training to pin an integration to a specific version. Idempotency-Key prevents double charging when the same request is sent again.
Model: nuliq · Bearer header required. /api/generate, a limited /api/chat, and /api/tags. One user message; no system messages, history, tools, or images. This is not a local Ollama model.
Track successful queries, input volume, and your balance.
Loading payment status…
One-time prepaid usage credit; no subscription or automatic renewal. Card details are processed by Stripe Checkout; NULIQ does not store your card number. Supported payment methods are shown at checkout. Terms · Refunds and cancellation · Delivery · Support
| Date | Model | Input | Output | Cost |
|---|
No queries yet. Get your first decision in Playground.