OpenAI has officially launched GPT-6.1 Terra, its latest artificial intelligence model designed to significantly enhance how AI interacts with computer interfaces. This upgraded model improves visual understanding, coding, and writing capabilities, making it a powerful tool for automating a wide range of professional tasks across various applications.
Revolutionizing Computer Interaction
GPT-6.1 Terra introduces advanced computer-use capabilities, allowing the AI to interpret on-screen information and execute multi-step tasks. This means the model can go beyond generating text; it can actively engage with software interfaces, provided it has the necessary tools and permissions. Potential applications include drafting proposals, interpreting complex technical drawings, accurately filling out forms, and even editing videos.
For businesses, this translates into potential efficiencies in operations such as processing incoming invoices and automatically entering details into accounting software, or updating records across different platforms. The model's ability to understand visual data is a key differentiator, enabling it to work more intuitively with graphical user interfaces.
Enhanced Performance and Efficiency
OpenAI highlights substantial performance improvements in GPT-6.1 Terra compared to its predecessor, GPT-5.6 Terra. In tests on the Agents’ Last Exam, a benchmark for professional task performance, GPT-6.1 Terra scored 15 percentage points higher at medium reasoning effort. Furthermore, it achieved this at approximately one-third the cost per task during the evaluation.
The new model also shows significant strides in coding proficiency. On Terminal-Bench 4.0, a test for command-line task resolution, GPT-6.1 Terra scored 50.3% at maximum reasoning effort, a notable increase from GPT-5.6 Terra’s 21.5%. These figures, reported by OpenAI, suggest a more capable and efficient AI agent, though actual performance will vary based on specific tasks and environments.
Significant API Price Reduction
A major draw for developers and businesses is the substantial reduction in API pricing. OpenAI has cut the standard API costs for GPT-6.1 Terra by 50% compared to GPT-5.6 Terra and GPT-6.1 Sol. The new pricing structure is:
- Input: $1 per million tokens
- Cached input: $0.05 per million tokens
- Output: $5 per million tokens
Tokens are small units of text processed by the AI model. This aggressive pricing strategy aims to make GPT-6.1 Terra more accessible and cost-effective for developers running high-volume applications, potentially lowering the overall cost of integrating advanced AI capabilities.
Availability and Future Prospects
GPT-6.1 Terra is currently accessible to Plus, Pro, Enterprise, and Edu users through ChatGPT Work and Codex. Developers can integrate the model into their applications via the OpenAI API, using the model name gpt-6.1-terra. While the model offers powerful automation capabilities, OpenAI advises users to review critical actions to ensure accuracy and prevent errors.
The combination of improved performance, advanced computer-use features, and significantly reduced API costs positions GPT-6.1 Terra as a compelling tool for those looking to leverage artificial intelligence for enhanced productivity and automation.