'daybreak-red-latest,' described as 'An alias for advanced cybersecurity models for authorized vulnerability research and security testing,' was removed and replaced by 'gpt-daybreak-red-latest' with identical notes. 'daybreak-blue-latest,' described as 'An alias for flagship general-purpose models with safeguards for defensive cybersecurity work,' was similarly replaced by 'gpt-daybreak-blue-latest.' Functional descriptions are unchanged; only the model slug prefix differs. Adding the 'gpt-' prefix places these specialized security aliases within the main GPT naming scheme, which may reflect an intent to surface them more visibly in the standard model catalog or to align them with GPT-branded API access tiers.
Every field of each cited change, before and after.
| Field | Before | After |
|---|---|---|
| cached_input_per_mtok | ∅ | ∅ |
| context_window | ∅ | ∅ |
| currency | USD | ∅ |
| effective_from | ∅ | ∅ |
| effective_until | ∅ | ∅ |
| input_per_mtok | ∅ | ∅ |
| model_slug | daybreak-red-latest | ∅ |
| notes | An alias for advanced cybersecurity models for authorized vulnerability research and security testing. | ∅ |
| output_per_mtok | ∅ | ∅ |
| provider | openai | ∅ |
| tier | ∅ | ∅ |
| Field | Before | After |
|---|---|---|
| cached_input_per_mtok | ∅ | ∅ |
| context_window | ∅ | ∅ |
| currency | ∅ | USD |
| effective_from | ∅ | ∅ |
| effective_until | ∅ | ∅ |
| input_per_mtok | ∅ | ∅ |
| model_slug | ∅ | gpt-daybreak-red-latest |
| notes | ∅ | An alias for advanced cybersecurity models for authorized vulnerability research and security testing. |
| output_per_mtok | ∅ | ∅ |
| provider | ∅ | openai |
| tier | ∅ | ∅ |
| Field | Before | After |
|---|---|---|
| cached_input_per_mtok | ∅ | ∅ |
| context_window | ∅ | ∅ |
| currency | ∅ | USD |
| effective_from | ∅ | ∅ |
| effective_until | ∅ | ∅ |
| input_per_mtok | ∅ | ∅ |
| model_slug | ∅ | gpt-daybreak-blue-latest |
| notes | ∅ | An alias for flagship general-purpose models with safeguards for defensive cybersecurity work. |
| output_per_mtok | ∅ | ∅ |
| provider | ∅ | openai |
| tier | ∅ | ∅ |
| Field | Before | After |
|---|---|---|
| cached_input_per_mtok | ∅ | ∅ |
| context_window | ∅ | ∅ |
| currency | USD | ∅ |
| effective_from | ∅ | ∅ |
| effective_until | ∅ | ∅ |
| input_per_mtok | ∅ | ∅ |
| model_slug | daybreak-blue-latest | ∅ |
| notes | An alias for frontier general-purpose models with safeguards for defensive cybersecurity work. | ∅ |
| output_per_mtok | ∅ | ∅ |
| provider | openai | ∅ |
| tier | ∅ | ∅ |
Data from the last 90 days is CC BY 4.0. Older data is CC BY-NC 4.0, except what we published before 2026-11-04, which stays CC BY 4.0. How you can reuse it.