Release
OpenAI presents GPT-6.1 Terra, a model designed for detailed visual reasoning and precise computer use.
According to OpenAI, the model reaches the Critical threshold in cybersecurity of its Preparedness Framework.
The protections developed for GPT-6 Astra are maintained and tests show strong resistance to prompt-injection.
On Terminal-Bench 4.0 at maximum reasoning effort, 50.3% of tasks are solved versus 21.5% for GPT-5.6 Terra.
Standard API prices are 50% lower than those of GPT-5.6 Terra and GPT-6.1 Sol.