Model Cards / Google DeepMind

Gemini 3 Pro Model Card

model card2,243 words·10 min read·Jul 27, 2026·Source
Version History
Chaptered summary is still being generated for this document. Showing a heuristic brief in the meantime.
Summary
2,243-word document condensed to 64 words. Google DeepMind · Jul 27, 2026
TL;DR

the Google DeepMind site for a comprehensive list of model cards. This model card includes more essential information about the Gemini 3 family of models than previous model

Top benchmarks
BenchmarkVariantScore
key skills benchmarkv1_hard, solve_rate91.7%
Misalignmentsituational_awareness, solve_rate27.3%
Misalignmentstealth, solve_rate25.0%
Tone7.9%
Unjustified-refusals3.7%
Image to Text Safety3.1%
Multilingual SafetyAverage20.0%
key skills benchmarkv2, solve_rate0.0%

Showing top 8 of 10. See full list below.

Deployment scope
  • available via Notebook LM.

Every italicized passage is a verbatim substring of the source document (checked deterministically after extraction). Field selection is heuristic — some quotes may lack surrounding context and some claims may be absent if no matching pattern appeared. For citation, open the source: original model card · source SHA 108e93b43e9d · version dated Jul 27, 2026.

Extracted Evaluations(10 results)

Sort by:0/10 rows fully reproducible (0%)
BenchmarkCategoryStateScoreSetupSource
key skills benchmark/ v1_hard
otherscored
91.7
solve rate
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
Misalignment/ situational_awareness
otherscored
27.3
solve rate
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
Misalignment/ stealth
otherscored
25.0
solve rate
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
otherscored
7.9
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
otherscored
3.7
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
otherscored
3.1
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
otherscored
0.2
Averagemissing: shot countmissing: methodmissing: training state
self-reported
key skills benchmark/ v2
otherscored
0.0
solve rate
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
otherscored
-10.4
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported
othermentioned
missing: shot countmissing: methodmissing: languagemissing: training state
self-reported