Model profile
Gemini 3.8 Flash Cyber
Evidence summary
Safety. Gemini 3.8 Flash Cyber does not meet the evidence gate for a Safety rank. Published Safety evidence spans 1 eval lineages and 1 of 7 components. Its strongest relative result is Google Gemini 3.8 launch — Gray Swan indirect prompt injection k=15 (attack_success_probability_k15_pct, #3 of 15); its weakest is Google Gemini 3.8 launch — Gray Swan indirect prompt injection k=15 (attack_success_probability_k15_pct, #3 of 15).
Freedom. Gemini 3.8 Flash Cyber does not meet the evidence gate for a Freedom rank.
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Official launch and evaluation methodology ↗Exact model document · google · first party
Safety evals
Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| Google Gemini 3.8 launch — Gray Swan indirect prompt injection k=15attack_success_probability_k15_pct | #3 / 15 | ↓6 | Source ↗official |
Freedom evals
No published sub-eval result contributes to this model’s Freedom profile.