Skip to main content

Best Model

Qwen 3.6 Plus (qwen-3.6-plus) — Best overall quality for varied automation tasks. ~$0.75 per 1K requests. For high-volume, low-cost automation: Gemini 2.5 Flash (gemini-2.5-flash) or Qwen 3 32B (qwen-3-32b) depending on context needs.

Python — Email Automation

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-your-key")

def draft_reply(email_body: str) -> str:
    response = client.chat.completions.create(
        model="qwen-3.6-plus",
        messages=[
            {"role": "system", "content": "Draft a professional reply to this email. Be concise and friendly."},
            {"role": "user", "content": email_body}
        ]
    )
    return response.choices[0].message.content

# Process incoming emails
emails = ["Can we reschedule our meeting?", "Please send the Q2 report"]
for email in emails:
    print(draft_reply(email))

Python — Batch Data Processing

import json
from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-your-key")

def classify_ticket(ticket: str) -> dict:
    response = client.chat.completions.create(
        model="qwen-3-32b",  # fast + structured classification
        messages=[
            {"role": "system", "content": "Classify this support ticket. Return JSON: {\"category\": \"...\", \"priority\": \"low|medium|high\", \"summary\": \"...\"}"},
            {"role": "user", "content": ticket}
        ],
        response_format={"type": "json_object"}
    )
    return json.loads(response.choices[0].message.content)

tickets = [
    "My payment failed and I can't access my account",
    "How do I change my email address?",
    "Your API has been down for 2 hours, this is critical"
]

for ticket in tickets:
    result = classify_ticket(ticket)
    print(f"[{result['priority']}] {result['category']}: {result['summary']}")

JavaScript — Scheduled Reports

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://kymaapi.com/v1",
  apiKey: process.env.KYMA_API_KEY,
});

async function generateReport(data: string): Promise<string> {
  const response = await client.chat.completions.create({
    model: "qwen-3.6-plus",
    messages: [
      { role: "system", content: "Analyze this data and write a brief executive summary with key insights and action items." },
      { role: "user", content: data },
    ],
  });
  return response.choices[0].message.content!;
}

// Run daily via cron
const metrics = "Users: 1,200 (+15%), Revenue: $5,400 (+8%), Churn: 2.1% (-0.3%)";
console.log(await generateReport(metrics));

Tips

  • Use qwen-3-32b for high-volume structured tasks (classification, extraction)
  • Use qwen-3.6-plus when quality matters (emails, reports, summaries)
  • Add response_format: {"type": "json_object"} for structured output
  • Batch requests where possible to reduce latency overhead

Cost Estimate

WorkflowVolumeModelDaily Cost
Email drafts50/dayqwen-3.6-plus~$0.04
Ticket classification500/dayqwen-3-32b~$0.18
Daily reports5/dayqwen-3.6-plus~$0.004

Next Steps