Gemini 3.1 Pro
16 values about Gemini 3.1 Pro from 2 documents, in 5 metric families. In each family, values from its own document come first.
- Developer
- Google DeepMind
- Release date
- 19 Feb 2026The date of its first card in the dataset, the Gemini 3.1 Pro Model Card.
- Availability
- Public
- Its own document
- Gemini 3.1 Pro Model Card · 19 Feb 2026
- Also reported in
- Gemini 3.7 Flash Frontier Safety Framework Report · 13 Aug 2026
Data as of 26 Sep 2026 · Dataset v0.1 · Methodology v0.1 · Changelog
Values by metric family
Each value as the document printed it. Select a value for its source, its checks and its history. Values from other documents are listed after the model’s own, each marked as the first report of that measure or as a restatement.
KeyVerifiedUnverifiedDisputedCorrected read off a figure or stated in wordsA value opens its source and history.
M1 · Evaluation awareness 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGemini 3.1 Pro Model Card · 19 Feb 2026 | ||
| Situational awareness challengesSuccess rate on three hardest challengesCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
M3 · Sabotage and sandbagging 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGemini 3.1 Pro Model Card · 19 Feb 2026 | ||
| Stealth challengesComparison with Gemini 3 ProCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
M4 · Misalignment audits 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGemini 3.1 Pro Model Card · 19 Feb 2026 | ||
| Frontier Safety Framework determination (misalignment)Instrumental reasoningCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
M7 · Harmful compliance and over-refusal 6 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGemini 3.1 Pro Model Card · 19 Feb 2026 | ||
| Child safety launch thresholdsLaunch thresholds met | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Image to Text SafetyImage-to-text policy-violation rate deltaCondition: vs Gemini 3 Pro | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Multilingual SafetyMultilingual policy-violation rate deltaCondition: vs Gemini 3 Pro | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Text to Text SafetyPolicy-violation rate deltaCondition: vs Gemini 3 Pro | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| ToneObjective tone of refusals deltaCondition: vs Gemini 3 Pro | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Unjustified refusalsUnjustified refusals deltaCondition: vs Gemini 3 Pro | Gemini 3.1 Pro Model Card19 Feb 2026 | |
M10 · Dangerous capabilities and risk determinations 7 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGemini 3.1 Pro Model Card · 19 Feb 2026 | ||
| Frontier Safety Framework determinationCBRN Uplift Level 1Condition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Frontier Safety Framework determinationCyber Uplift Level 1 | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Frontier Safety Framework determinationHarmful manipulationCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Frontier Safety Framework determinationML R&DCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| Harmful manipulation efficacy studyMax odds ratio vs non-AI baselineCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| RE-BenchAverage normalised scoreCondition: Deep Think | Gemini 3.1 Pro Model Card19 Feb 2026 | |
| From other documents | ||
| GRB internal research-engineering benchmarkAverage pass@1Gemini 3.7 Flash Frontier Safety Framework Report13 Aug 2026First reported | Gemini 3.7 Flash Frontier Safety Framework Report13 Aug 2026First reported | |
Risk determinations
The developer's formal decisions about Gemini 3.1 Pro under its framework, as printed. Levels from different frameworks do not map onto one another.
| Domain | Level as printed | Framework and document |
|---|---|---|
| Bio/chem | not_reached code, not the printed wording Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 | Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 |
| Cyber | alert_threshold_reached code, not the printed wording Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 | Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 |
| AI R&D / autonomy | not_reached code, not the printed wording Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 | Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 |
| Manipulation | not_reached code, not the printed wording Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 | Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 |
| Misalignment | not_reached code, not the printed wording Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 | Frontier Safety Framework Gemini 3.1 Pro Model Card · 19 Feb 2026 |
| Other | met code, not the printed wording Internal launch thresholds Gemini 3.1 Pro Model Card · 19 Feb 2026 | Internal launch thresholds Gemini 3.1 Pro Model Card · 19 Feb 2026 |
“Code, not the printed wording”: the version 0 file stored a code for this level rather than the words the document printed. The wording will be read again from the source.