Google DeepMind · Gemma 3

Gemma 3 27B IT

READER SUMMARY

The largest instruction-tuned Gemma 3 checkpoint with multimodal reasoning, function calling and long context on a single accelerator.

DOOM SCORE62.6out of 100model risk profile, not the overall index
CURRENT ASSESSMENT · REVISION 1

Why this model scores 62.6

The strongest Gemma 3 tier combines capable multimodal generation and tool-oriented outputs with unusually accessible hardware requirements. Open weights materially elevate misuse exposure and control difficulty.

Capability66
Autonomy22
Deployment98
Misuse potential62
Control difficulty69
0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

MODEL-ATTRIBUTED EVIDENCE

News tied to Gemma 3 27B IT

The model score of 62.6 rates this model's risk profile. The overall Doom Index of 63.5 measures the recalibrated complete evidence corpus. These values answer different questions.

NET MODEL-ATTRIBUTED INDEX CONTRIBUTION+0.00

Each article's leave-one-out Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores are recalculated from their technical profile and related news, but never feed back into the overall index.

AUDIT TRAIL

Model score history

  1. R1
    Doom Score 62.6

    Initial exact profile for the separately released 27B instruction checkpoint.

    12 Aug 2026