Back to Models
Google Deep MindConditional Open

Gemma 3-270M

Gemma 3-270M is a foundation model developed by Google DeepMind. Despite its compact size, it is equipped with a broad 128K context window.

Parameters

2.0B

Context Window

128K

License

Gemma License

Release Date

2025-08-14

API Pricing

API pricing for this model is not yet available

Strengths

  • Lightweight and efficient model size
  • Supports long 128K context
  • Developed by Google DeepMind

Weaknesses

  • Limited parameters, unsuitable for complex tasks
  • Limited knowledge compared to larger models
  • Lacks advanced reasoning capabilities

Use Cases

  • Low-latency operation on edge devices
  • Small-scale text generation tasks
  • Efficient document context processing

Deep Analysis

Parameters

270M

170M embedding + 100M transformer blocks

Vocabulary

256K tokens

Large vocabulary for handling rare/specific tokens

Release Date

August 14, 2025

Part of Gemma 3 family expansion

Battery Usage

0.75% per 25 conversations

Tested on Pixel 9 Pro SoC with INT4 quantization

Quantization

INT4 QAT

Quantization-Aware Trained checkpoints available

License

Gemma Terms of Use

Gemma Terms of Use

IFEval

Best-in-class for size

New level of instruction-following performance for 270M class

Strengths

  • Extremely energy efficient: 0.75% battery for 25 conversations on Pixel 9 Pro
  • Strong instruction-following out of the box (best IFEval for its size class)
  • 256K vocabulary handles rare and domain-specific tokens well
  • Production-ready INT4 QAT quantization for edge deployment
  • Excellent base for task-specific fine-tuning

Weaknesses

  • Not designed for complex conversational use cases
  • Very limited general knowledge and reasoning capability
  • Text-only (no multimodal support)
  • Small context window compared to larger models
  • Requires fine-tuning for production quality on specific tasks

Competitor Comparison

ModelArenaSWEGPQAPrice
Gemma 3 270MN/AN/AN/AFree (open weights)
Phi-3.5 Mini (3.8B)~1100N/A~55%Free (open weights)
SmolLM2 (1.7B)N/AN/AN/AFree (open weights)
Qwen2.5 (0.5B)N/AN/AN/AFree (open weights)

Gemma 3 270M is Google's most compact open model, designed for task-specific fine-tuning with strong instruction-following capabilities. With only 270M parameters and extreme energy efficiency (0.75% battery per 25 conversations on Pixel 9 Pro), it targets on-device AI, IoT, and specialized classification/extraction tasks.

Analysis generated: 2026-05-24