Gemma-3
google/gemma-3-27b-it- Vendor
- Parameters
- 27B
- Size
- 50.3 GB
$ brew install mirai$ mirai --model google/gemma-3-27b-itIntegrate with SDK
https://github.com/trymirai/uzu-swift
| 1 | import Foundation |
| 2 | import Uzu |
| 3 | |
| 4 | public func runChat() async throws { |
| 5 | let engineConfig = EngineConfig.create() |
| 6 | let engine = try await Engine.create(config: engineConfig) |
| 7 | |
| 8 | guard let model = try await engine.model(identifier: "alibaba:qwen3.5:0.8b:mirai:mirai-m:4") else { |
| 9 | return |
| 10 | } |
| 11 | for try await update in try await engine.download(model: model).iterator() { |
| 12 | print(String(format: "\r\u{001B}[2KDownload progress: %.2f%%", update.progress() * 100), terminator: "") |
| 13 | fflush(stdout) |
| 14 | } |
| 15 | print() |
| 16 | |
| 17 | let messages = [ |
| 18 | ChatMessage.system().withText(text: "You are a helpful assistant"), |
| 19 | ChatMessage.user().withText(text: "Tell me a short, funny story about a robot") |
| 20 | ] |
| 21 | let session = try await engine.chat(model: model, config: .create()) |
| 22 | let stream = await session.replyWithStream(input: messages, config: .create()) |
| 23 | var message: ChatMessage? = nil |
| 24 | for try await update in stream.iterator() { |
| 25 | switch update { |
| 26 | case .replies(let replies): |
| 27 | let reply = replies.last |
| 28 | message = reply?.message |
| 29 | print("Generated tokens: \(reply?.stats.tokensCountOutput ?? 0)") |
| 30 | case .error(let error): |
| 31 | print("Error: \(error)") |
| 32 | } |
| 33 | } |
| 34 | print("Reasoning: \(message?.reasoning() ?? "empty")") |
| 35 | print("Text: \(message?.text() ?? "empty")") |
| 36 | } |
| 37 | |
Details
Gemma 3 is Google's family of lightweight, state-of-the-art open models, built from the same research and technology behind the Gemini models. The 27B instruction-tuned (IT) variant is the largest in the Gemma 3 lineup, offering strong performance across a wide range of tasks while remaining deployable on accessible hardware.
Multimodal & Multilingual
Gemma 3 27B IT is a multimodal model that accepts both text and image inputs and generates text output. Images are normalized to 896×896 resolution, and the model can produce responses of up to 8,192 tokens. It supports a generous 128K context window, making it well-suited for tasks that require processing lengthy documents or extended conversations. Multilingual capability spans over 140 languages, significantly broadening its applicability across global use cases.
Key Capabilities
- Question answering — grounded responses across diverse domains
- Summarization — distilling long-form content into concise outputs
- Reasoning — multi-step inference over text and visual inputs
- Image understanding — interpreting and describing visual content alongside text
Architecture & Deployment
As an instruction-tuned model, Gemma 3 27B IT has been fine-tuned to follow natural language instructions, making it practical for conversational AI, assistants, and agentic workflows. Despite its 27 billion parameter scale, it is designed to run in resource-conscious environments — from custom cloud infrastructure down to high-end desktop setups — helping democratize access to capable open-weight models.
Provenance
Developed by Google and released with open weights, Gemma 3 27B IT carries forward Google's commitment to making frontier-grade AI research broadly available. Both pre-trained and instruction-tuned variants are offered across multiple sizes, with this 27B IT checkpoint representing the flagship configuration for users seeking maximum capability.