GPT-5.2-Codex
13 values about GPT-5.2-Codex from 3 documents, in 5 metric families. In each family, values from its own document come first.
- Developer
- OpenAI
- Release date
- 18 Dec 2025The date of its first card in the dataset, the GPT-5.2-Codex System Card Addendum.
- Availability
- Public
- Its own document
- GPT-5.2-Codex System Card Addendum · 18 Dec 2025
- Also reported in
- GPT-5.3-Codex System Card · 5 Feb 2026
- GPT-5.4 Thinking System Card · 5 Mar 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.
M3 · Sabotage and sandbagging 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From other documents | ||
| Sabotage capabilityMean best-of-10 scoreRun by Apollo ResearchGPT-5.3-Codex System Card5 Feb 2026First reported | GPT-5.3-Codex System Card5 Feb 2026First reported | |
M4 · Misalignment audits 2 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGPT-5.2-Codex System Card Addendum · 18 Dec 2025 | ||
| Destructive action avoidanceAvoidanceRestated later: compare with the later value | GPT-5.2-Codex System Card Addendum18 Dec 2025Restated later: compare with the later value | |
| From other documents | ||
| Destructive action avoidanceAvoidanceGPT-5.3-Codex System Card5 Feb 2026Restated: compare with the earlier value | GPT-5.3-Codex System Card5 Feb 2026Restated: compare with the earlier value | |
M7 · Harmful compliance and over-refusal 4 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGPT-5.2-Codex System Card Addendum · 18 Dec 2025 | ||
| Cyber safetyProduction data | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
| Cyber safetySynthetic data | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
| Production BenchmarksEmotional reliance | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
| Production BenchmarksMental health | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
M8 · Jailbreak robustness 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGPT-5.2-Codex System Card Addendum · 18 Dec 2025 | ||
| StrongRejectnot_unsafe (aggregate) | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
M10 · Dangerous capabilities and risk determinations 5 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentGPT-5.2-Codex System Card Addendum · 18 Dec 2025 | ||
| Preparedness Framework determinationAI self-improvement | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
| Preparedness Framework determinationBiological and chemical | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
| Preparedness Framework determinationCybersecurity | GPT-5.2-Codex System Card Addendum18 Dec 2025 | |
| From other documents | ||
| Cyber RangeCombined pass rateGPT-5.3-Codex System Card5 Feb 2026First reportedRestated later: compare with the later value | GPT-5.3-Codex System Card5 Feb 2026First reportedRestated later: compare with the later value | |
| Cyber RangeCombined pass rateGPT-5.4 Thinking System Card5 Mar 2026Restated: compare with the earlier value | GPT-5.4 Thinking System Card5 Mar 2026Restated: compare with the earlier value | |
Restated in later documents
A later document reported values about GPT-5.2-Codex again. Each pair is shown side by side: both values are kept with their own documents, and the later one does not replace the earlier.
| Earlier value | Later value |
|---|---|
| M4 · Misalignment audits | |
| Destructive action avoidanceAvoidance | |
| GPT-5.2-Codex System Card Addendum18 Dec 2025 · its own document | GPT-5.3-Codex System Card5 Feb 2026No reason stated |
| M10 · Dangerous capabilities and risk determinations | |
| Cyber RangeCombined pass rate | |
| GPT-5.3-Codex System Card5 Feb 2026 · first reported | GPT-5.4 Thinking System Card5 Mar 2026No reason stated |
Risk determinations
The developer's formal decisions about GPT-5.2-Codex under its framework, as printed. Levels from different frameworks do not map onto one another.
| Domain | Level as printed | Framework and document |
|---|---|---|
| Bio/chem | High (treated as) Preparedness Framework GPT-5.2-Codex System Card Addendum · 18 Dec 2025 | Preparedness Framework GPT-5.2-Codex System Card Addendum · 18 Dec 2025 |
| Cyber | Below High Preparedness Framework GPT-5.2-Codex System Card Addendum · 18 Dec 2025 | Preparedness Framework GPT-5.2-Codex System Card Addendum · 18 Dec 2025 |
| AI R&D / autonomy | Below High Preparedness Framework GPT-5.2-Codex System Card Addendum · 18 Dec 2025 | Preparedness Framework GPT-5.2-Codex System Card Addendum · 18 Dec 2025 |