Skip to main content

Gemma 4 Inference (Yamify)

Yamify hosts Gemma 4 so you can ship chatbots and AI features without running model infra.

Endpoint​

Yamify provides an OpenAI-compatible Chat Completions endpoint:

  • Base URL: https://www.yamify.co/api/inference/v1
  • Chat completions: POST /chat/completions
  • Models: GET /models

Auth (MVP)​

Create an API key in Yamify → Dashboard → Settings → API keys.

Then send it as:

  • Authorization: Bearer <YOUR_KEY>

Keys are shown once when created (store it in a password manager).

Quick test (curl)​

curl -sS https://www.yamify.co/api/inference/v1/chat/completions \\
  -H "content-type: application/json" \\
  -H "authorization: Bearer $YAMIFY_API_KEY" \\
  -d '{
    "model": "yamify/gemma4",
    "stream": false,
    "messages": [
      { "role": "system", "content": "You are a helpful assistant." },
      { "role": "user", "content": "Write a 1-paragraph README for a TODO app." }
    ],
    "temperature": 0.2
  }' | jq

If you get a warming up error, wait ~2 minutes and retry (first-time model pull).

Gemma 4 can be slow to answer on CPU. For the best UX (and to avoid proxy timeouts), use streaming:

curl -sS --no-buffer https://www.yamify.co/api/inference/v1/chat/completions \\
  -H "content-type: application/json" \\
  -H "authorization: Bearer $YAMIFY_API_KEY" \\
  -d '{
    "model": "yamify/gemma4",
    "stream": true,
    "messages": [
      { "role": "user", "content": "Count from 1 to 5." }
    ],
    "temperature": 0.2
  }'

This returns an OpenAI-style SSE stream (data: {...}\n\n), ending with data: [DONE].

Node.js example (OpenAI SDK)​

npm i openai
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.YAMIFY_API_KEY,
  baseURL: "https://www.yamify.co/api/inference/v1",
});

const completion = await client.chat.completions.create({
  model: "yamify/gemma4",
  messages: [{ role: "user", content: "Say hi from Gemma 4." }],
  stream: false,
});

console.log(completion.choices[0].message.content);

Deploy a sample chatbot on Yamify (custom app)​

  1. Create a new Next.js app:
npx create-next-app@latest gemma4-chatbot --ts --eslint
cd gemma4-chatbot
npm i openai
  1. Add app/api/chat/route.ts:
import OpenAI from "openai";

export async function POST(req: Request) {
  const { messages } = await req.json();
  const client = new OpenAI({
    apiKey: process.env.YAMIFY_API_KEY,
    baseURL: "https://www.yamify.co/api/inference/v1",
  });

  const completion = await client.chat.completions.create({
    model: "yamify/gemma4",
    messages,
    temperature: 0.2,
  });

  return Response.json({ reply: completion.choices[0].message.content });
}
  1. Deploy it with Yamify:
    • Go to Dashboard → Deploy from GitHub or ZIP
    • Push this repo to GitHub and paste the URL
    • Set env var YAMIFY_API_KEY in Yamify for the app (same key you use locally)