Gemma 3 27B
Language modelPrevious generationOpen weightsCard summarizedA family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models
Previous generation; open weights still downloadable● Re-checked 12 Oct 2026 · 8 facts · 0 corrections · 1 unconfirmedSource: Gemma 3 model card
Use it via API
Where to call it
- Self-host (weights)
google/gemma-3-27b-it - Google Vertex AI
Gemma 3 27B
Good for
- Text generation
- Chatbots and conversational AI
- Text summarization
Overview
Gemma 3 27B
A family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models
- Cost
- $0.08 host
- Context
- 128K
- Cheapest host
- $0.08 / $0.16
- Max output
- 128K
- Knows up to
- August 2024
- Released
- 10 Mar 2025
Lightweight open multimodal models built from the same research as Gemini. They take text and images and return text, with a 128K context window and support for over 140 languages, and are small enough to run on laptops, desktops or your own cloud. Model information
gemma-3-27b-it- Input / output per 1M
- $0.08 / $0.16 via DeepInfra
- Context window
- 128K
- Max output
- 128K
- Knows up to
- August 2024
- Reasoning
- —
- Weights
- Open · Gemma Terms of Use
About this card. A model card is the maker's own account of a model: what it's for, how it was trained and tested, and where it falls short. This page condenses Google DeepMind's Gemma 3 model card. Every claim carries its source page.