Google πŸ‡ΊπŸ‡Έ Β· Gemini 2.5

Gemini 2.5 Computer Use

A proprietary Gemini 2.5 Pro-derived model specialized for browser and mobile interface control through screenshot-conditioned action loops.

DOOM SCORE79.7out of 100model risk profile, not the overall index
0 comments Β· 0 votesOpen discussion

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

No comments yet. Start the discussion.

CURRENT ASSESSMENT Β· REVISION 1

Why this model scores 79.7

Gemini 2.5 Pro reasoning plus direct clicking, typing, scrolling, form completion, and logged-in interaction create high practical autonomy. Public API preview and confirmed production use establish meaningful but not universal deployment. Prompt injection, scams, unexpected behavior, and high-risk action pathways raise misuse and control difficulty, moderated by client-side execution, per-step safety review, confirmations, system instructions, and the preview restriction.

Capability86
Autonomy88
Deployment78
Misuse potential79
Control difficulty63
MODEL-ATTRIBUTED EVIDENCE

News tied to Gemini 2.5 Computer Use

The model score of 79.7 rates this model's risk profile. The overall Doom Index of 61.0 measures the complete temporally weighted evidence record. These values answer different questions.

NET MODEL-ATTRIBUTED INDEX CONTRIBUTION+0.36

Each article's current 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 use a bounded temporal offset around their technical profile, but never feed back into the overall index.

Autonomy TOWARD

Gemini 2.5 Computer Use gives agents direct browser control

Google released Gemini 2.5 Computer Use in public API preview for agents that click, type, scroll, and operate behind logins. Google also documented per-step safety checks, developer-defined confirmations, and restrictions for high-risk actions, while confirming production use in UI testing and agentic Search features.

Full item contribution
+0.36
Gemini 2.5 Computer Use equal share
+0.36
Read assessment β†’
AUDIT TRAIL

Model score history

  1. R1
    Doom Score 79.7

    New exact public model verified from its dated launch page and model card.

    14 Aug 2026