Uzbek LLM and OpenAI-compatible API

Chat with the NeuronAI assistant, meet MustaqiLLM, our Uzbek language model built from scratch, and connect your product through an OpenAI-compatible API.

What is an LLM? Read the guide

Try Large Language Model

Use the NeuronAI widget to test how one assistant can handle quick tasks, follow-up questions, and structured answers in a natural conversation flow.

MustaqiLLM, our Uzbek model built from scratch

MustaqiLLM is NeuronAI’s own language model for Uzbek. We designed and trained it from scratch instead of fine-tuning a foreign model, and its weights are available on Hugging Face.

  • 5B parameters
  • 40B token multilingual dataset
  • 48K token BPE tokenizer

Model weights on Hugging Face Read the announcement

OpenAI-compatible LLM API

Point an OpenAI SDK at NeuronAI: requests and answers keep the chat completions format, so moving an application over takes a new base URL and API key. Add stream: true to receive the answer as it is written.

Base URL
https://my.neuronai.uz/api/v1
Authentication
Authorization: Bearer <your-api-key>
Models
GET /models
Chat
POST /chat/completions

LLM API docs

Python · OpenAI SDK
from openai import OpenAI

client = OpenAI(base_url="https://my.neuronai.uz/api/v1", api_key="<your-api-key>")

completion = client.chat.completions.create(
    model="gemma-4-26b-a4b",
    messages=[{"role": "user", "content": "Salom! Bugun ob-havo qanday?"}],
)

print(completion.choices[0].message.content)

Use any model that GET /models lists for your account.

Simple, transparent pricing

Pay only for what you use — no monthly fees, ever.

Language Model (LLM)

Input

1.5 UZS / 1K tokens

Output

6 UZS / 1K tokens

  • OpenAI-compatible Chat Completions API
  • Streaming and function calling
  • Long-context conversations
  • Simple REST API

See all prices

Frequently Asked Questions

Tell us what you are building

We will help you choose the right NeuronAI stack for your workload.

Contact sales