Unlock the Hive Mind.

SharX provides high-performance LLM inference by aggregating idle GPU power globally. Get DeepSeek-R1, Llama-3, and Mistral speeds at 80% lower cost than Big Tech clouds.

API Base URL

https://apillm.semburat.online/v1

Use this as your base_url in OpenAI SDKs.

Blazing Fast

Low-latency inference running on H100 & A100 clusters nearby.

Drop-in Ready

100% compatible with OpenAI libraries. Just change the URL.

Private & Secure

No training on your data. Ephemeral processing nodes only.

Cost Effective

Pay only for computed tokens. No idle server fees.

Authentication

SharX API uses Bearer Token authentication.

Get API Key →

HTTP HEADER

Authorization: Bearer sk-sharx-YOUR_SECRET_KEY

Content-Type: application/json

Tip: Your API key carries full privileges. Do not share it in client-side code (browsers/apps). Always route requests through your own backend server.

POST

/v1/chat/completions

Request

from openai import OpenAI

client = OpenAI(
    base_url="https://apillm.semburat.online/v1",
    api_key="sk-sharx-xxxxxxxx"
)

response = client.chat.completions.create(
    model="deepseek-r1:32b",
    messages=[{"role": "user", "content": "Hello!"}],
    temperature=0.7
)
print(response.choices[0].message.content)

curl https://apillm.semburat.online/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-sharx-xxxxxxxx" \
  -d '{
    "model": "deepseek-r1:32b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

import OpenAI from "openai";

const openai = new OpenAI({
    baseURL: 'https://apillm.semburat.online/v1',
    apiKey: 'sk-sharx-xxxxxxxx',
});

async function main() {
    const completion = await openai.chat.completions.create({
        messages: [{ role: 'user', content: 'Hello!' }],
        model: 'deepseek-r1:32b',
    });
    console.log(completion.choices[0].message.content);
}
main();

Response

JSON

{
  "id": "chatcmpl-sharx-123",
  "object": "chat.completion",
  "created": 1709123456,
  "model": "deepseek-r1:32b",
  "choices": [{
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "Hello! I am SharX AI."
    },
    "finish_reason": "stop"
  }],
  "usage": { "total_tokens": 35 }
}

GET

/v1/models

from openai import OpenAI
client = OpenAI(base_url="...", api_key="...")
print(client.models.list())

curl https://apillm.semburat.online/v1/models \
  -H "Authorization: Bearer sk-sharx-xxxxxxxx"

GET

/v1/balance

Request

import requests
res = requests.get("https://apillm.semburat.online/v1/balance", 
    headers={"Authorization": "Bearer ..."})
print(res.json())

curl https://apillm.semburat.online/v1/balance \
  -H "Authorization: Bearer sk-sharx-xxxxxxxx"

Response

JSON

{
  "object": "balance",
  "total_credits": 100.00,
  "used_credits": 12.50,
  "remaining_credits": 87.50,
  "currency": "USD",
  "status": "active"
}

POST

/v1/tokenize

Request

import requests
data = {"model": "deepseek-r1:32b", "content": "Count me!"}
res = requests.post(
    "https://apillm.semburat.online/v1/tokenize", 
    json=data, 
    headers={"Authorization": "Bearer ..."}
)
print(res.json())

curl https://apillm.semburat.online/v1/tokenize \
  -H "Authorization: Bearer sk-sharx-xxxxxxxx" \
  -d '{"model": "deepseek-r1:32b", "content": "Count me!"}'

Response

JSON

{
  "object": "list",
  "count": 3,
  "tokens": [1520, 220, 11],
  "model": "deepseek-r1:32b"
}