Ziwol API — Tutorial

Selamat Datang di Ziwol API 🚀

Ziwol API menggunakan standar /v1/messages yang kompatibel dengan format Anthropic. Anda bisa menggunakan API ini tanpa SDK tambahan, cukup gunakan HTTP request standar seperti fetch atau requests.

Semua request akan diproses oleh Sava AI — asisten cerdas buatan Ziwol.

Quick Start

// Buat class Ziwol sendiri tanpa SDK luar
class Ziwol {
  constructor({ apiKey, baseURL }) {
    this.apiKey = apiKey;
    this.baseURL = baseURL;
  }

  async messages({ model, messages, max_tokens, thinking }) {
    const response = await fetch(`${this.baseURL}/messages`, {
      method: 'POST',
      headers: {
        'Content-Type': 'application/json',
        'x-api-key': this.apiKey,
        'anthropic-version': '2023-06-01'
      },
      body: JSON.stringify({ model, messages, max_tokens, thinking })
    });
    
    const data = await response.json();
    if (response.ok && data.content) {
      return data;
    } else {
      throw new Error(data?.error?.message || 'API Error');
    }
  }
}

// Inisialisasi client Ziwol
const client = new Ziwol({
  apiKey: "sk-zw-4434684d218aa2ff288f94de8ee3502ae4ddff04e00b2eec",
  baseURL: "https://api.ziwol.com/v1",
});

// Eksekusi API
async function main() {
  try {
    const reply = await client.messages({
      model: "sava-flash",
      max_tokens: 1024,
      thinking: "false", // "false" = hemat token, "true" = AI mikir lebih dalam
      messages: [{ role: "user", content: "Hello!" }],
    });
    console.log(reply.content[0].text);
  } catch (e) {
    console.error("Error:", e.message);
  }
}

main();

# import requests (pip install requests)
import requests

API_KEY = "sk-zw-4434684d218aa2ff288f94de8ee3502ae4ddff04e00b2eec"
BASE_URL = "https://api.ziwol.com/v1"

payload = {
    "model": "sava-flash",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello!"}]
}

response = requests.post(
    f"{BASE_URL}/messages",
    headers={
        "Content-Type": "application/json",
        "x-api-key": API_KEY,
        "anthropic-version": "2023-06-01"
    },
    json=payload
)

data = response.json()

# Pengecekan respon sukses
if response.status_code == 200 and "content" in data:
    print(data["content"][0]["text"])
else:
    print(f"Status Code: {response.status_code}")
    print("Response API:", data)

curl -X POST https://api.ziwol.com/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-zw-4434684d218aa2ff288f94de8ee3502ae4ddff04e00b2eec" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "sava-flash",
    "max_tokens": 1024,
    "thinking": "false",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

sava-flash

Gratis 10.000 token/hari. Reset 24 jam. Cocok untuk eksperimen.

sava-2.0

Model canggih (PDRP). Bayar sesuai pakai/plan.

sava-omni

Pro Coding (GPT-5.6 Sol + Luna). Eksekusi kode kompleks.

💡 Tips Hemat Token

Set parameter thinking: "false" di request body kamu. Saat ini di-set false, AI akan melewati proses "thinking/enhancement" sehingga token yang dipakai lebih sedikit, namun jawaban tetap akurat berdasarkan prompt kamu. Gunakan true hanya untuk tugas coding berat yang butuh analisa dalam.