Gemma 3 4B

Function calling Vision Chat

Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3-12B is Google's latest open source model, successor to Gemma 2

What it's best for

Best for high-volume, price-sensitive work: classification, extraction, routing, tagging and short drafts, where you are making thousands of calls and every fraction of a penny per million tokens counts. It supports function calling, so it can drive tools and agent workflows.

Pricing

Prices updated daily — last generated 2026-07-08.

Billing is metered per request in GBP on the same monthly invoice as your apps — no subscription, no minimum. We list every model once, at the cheapest route we can serve it on; if we have to fail over to a more expensive route, we absorb the difference and your price does not change.

Where your data is processed

Requests to Gemma 3 4B are routed to a vetted partner provider and processed on that provider's infrastructure — they leave our infrastructure, and the UK-residency guarantee that applies to our UK-hosted models does not apply here. We hold the provider credentials server-side and route on cost and availability. If you need prompts that never leave our own infrastructure, use one of our UK-hosted models.

Call it in two minutes

The gateway is OpenAI-compatible: point your SDK or HTTP client at api.node.uk and use the model id gemma-3-4b-it. Credentials come from your portal, with £15 of free credit on signup.

curl https://api.node.uk/api/v1/models/gemma-3-4b-it/v1/chat/completions \
  -H "Authorization: Bearer $NODE_GATEWAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"model": "gemma-3-4b-it", "messages": [{"role": "user", "content": "Hello"}]}'

from openai import OpenAI

client = OpenAI(
    base_url="https://api.node.uk/api/v1/models/gemma-3-4b-it/v1",
    api_key=NODE_GATEWAY_TOKEN,
)

reply = client.chat.completions.create(
    model="gemma-3-4b-it",
    messages=[{"role": "user", "content": "Hello"}],
)

Related models

Browse the full model catalogue, read which model for which task, or see how the AI Gateway works.