Google: Gemma 4 26B A4B IT

stable

Gemma 4 26B A4B IT is available through Everstack's OpenAI-compatible gateway with routing, fallback, retries, and observability.

gemma-4-26b-a4b-itgemmareleased Apr 2, 2026

Provider Quick start

Context

262.1K

tokens

Max output

32.8K

tokens

Input

per 1M

Output

per 1M

Activity

Daily period usage

Aggregated managed Everstack traffic. Buckets below 5 tenants or 100 requests are withheld.

Providers

ProviderContextInput /MOutput /M

Google

262.1K--

Pricing

Item	Price	Unit
Input	-	per 1M tokens
Output	-	per 1M tokens

Performance

Reliability

Compare

Compare usage and performance

Place this model beside up to three other catalog models.

Capabilities

Modalities

Input

Output

Details

Model ID: gemma-4-26b-a4b-it
Family: gemma
Status: stable
Knowledge: -
Released: Apr 2, 2026

Gateway Config

API: 2024-01-01
Auth: GOOGLE_API_KEY
Discovery: static
Provider type: static

Quick Start

index.ts

import Everstack from "@everstack/node";const client = new Everstack({  apiKey: process.env.EVERSTACK_API_KEY!,  baseUrl: "https://{instance}.{region}.everstack.ai",});const completion = await client.chat.completions.create({  model: "@google/gemma-4-26b-a4b-it",  messages: [{ role: "user", content: "Explain vector cache routing." }],});console.log(completion.choices[0]?.message?.content);

Endpoint

Base URL

https://generativelanguage.googleapis.com/v1

Chat completions

https://generativelanguage.googleapis.com/v1/chat/completions

Model

gemma-4-26b-a4b-it

Auth env

GOOGLE_API_KEY

Gemma 4 26B A4B IT FAQ

What is the context window of Gemma 4 26B A4B IT?

Gemma 4 26B A4B IT accepts up to 262.1K tokens of context and returns up to 32.8K output tokens on Google.

What can Gemma 4 26B A4B IT do?

Gemma 4 26B A4B IT supports chat, function calling, image input and reasoning. It accepts text and image input.

How do I call Gemma 4 26B A4B IT through Everstack?

Point any OpenAI-compatible SDK at your Everstack gateway and set the model to @google/gemma-4-26b-a4b-it. Everstack adds retries, provider fallback, caching, per-tenant budgets and OpenTelemetry traces for every request.