Cyberax AI Playbook
cyberax.com
Model · Google · Text

Gemma 4 31B

Google's flagship open-weight Gemma 4. Natively multimodal (text + vision); supersedes Gemma 2 with significantly improved capabilities and a larger size.

Modality
Text
License
Gemma Terms of Use (Open weights)
Parameter size
31B
Released
April 2, 2026
Last verified
June 8, 2026
Runs locally
Yes
Also handles
vision

Strengths

  • Open weights with strong multimodal performance
  • Native vision capability — no separate vision variant needed
  • Backed by Google's training infrastructure

Weaknesses

  • Larger memory footprint than Gemma 2 9B
  • Gemma Terms restrict some use cases

Try it

WhereTypeNotes
Hugging Face weights Gated — accept Gemma Terms
Google AI Studio hosted-api Free tier

Version history

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

Change log

  • — Verified current vs Google docs. Corrected release date to 2026-04-02 (Gemma 4 launch per Google blog; prior 2026-05-07 was an HF lastModified timestamp).
  • — Initial entry. Sighted on Hugging Face (lastModified 2026-05-07).