Cyberax AI Playbook
cyberax.com
Model · Google · Text

Gemma 2 9B

Google's open-weight family. The 9B variant is the practical sweet spot — runs on a single GPU and outperforms older 13B-class models on most benchmarks.

Modality
Text
License
Gemma Terms of Use (Open weights)
Parameter size
9B
Context window
8,192 tokens
Released
June 27, 2024
Last verified
May 10, 2026
Runs locally
Yes

Strengths

  • Backed by Google's training infrastructure
  • Runs on a single consumer GPU
  • Strong instruction following for size

Weaknesses

  • 8k context window is small by 2026 standards
  • Gemma Terms restrict some use cases

Try it

WhereTypeNotes
Hugging Face weights Gated — accept Gemma Terms
Ollama local ollama run gemma2
Google AI Studio hosted-api Free tier

Version history

  1. Gemma 4 31B May 2026
  2. Gemma 2 9B Jun 2024 Current Deprecated

Change log

  • — Initial entry.