OpenAI Launches GPT-6.1 Terra Model Focused on Fast Computer Use
OpenAI has introduced GPT-6.1 Terra, an AI model designed for faster, lower-cost enterprise tasks, including coding, document drafting and desktop automation. It offers improved benchmark scores, reduced factual errors and enhanced security safeguards. The model is available to Plus, Pro, Enterprise and Edu subscribers, with API pricing cut by 50%.
Artificial intelligence research laboratory OpenAI has introduced GPT-6.1 Terra, a new AI model engineered for faster, cost-effective performance across routine enterprise tasks including document drafting, visual analysis, coding, and direct desktop application interaction. The release represents a major update over its predecessor, GPT-5.6 Terra, delivering enhanced accuracy and expanded computer-use capabilities.
Chief Executive Officer Sam Altman announced the deployment on social media platform X, highlighting the model's specialized proficiency in desktop environment navigation. As per a post by OpenAI, the company has cut standard API pricing by 50 percent compared to earlier variants, making high-volume workflow automation significantly more accessible for software developers. OpenAI Will Not Go Public in 2026; CEO Sam Altman Cites AI Safety Focus.
Benchmark Enhancements and Desktop Application Integration
GPT-6.1 Terra demonstrates substantial efficiency gains across visual reasoning and autonomous computer-use evaluations. On the Agents' Last Exam benchmark, the model scored 15 percentage points higher than GPT-5.6 Terra at medium reasoning effort while running at roughly one-third of the operational cost.
The model also completed tasks more than three times faster than higher-tier counterparts such as GPT-6 Astra and GPT-6.1 Sol. On the OSWorld 2.1 evaluation for desktop application navigation, GPT-6.1 Terra surpassed previous benchmarks, offering high token efficiency and up to a ninefold cost reduction per task compared to competing industry models.
Lower Factual Error Rates and Enhanced Security Safeguards
OpenAI reported significant progress in reducing model hallucinations and factual errors. Internal benchmark evaluations showed the factual error rate at low reasoning effort dropped to 16.7 percent from 53.6 percent in the previous generation, while maximum reasoning effort reduced errors to 4.3 percent. On Terminal-Bench 4.0 coding tasks, the model solved 50.3 percent of command-line challenges. OpenAI To Double Workforce to 8,000 Employees by Late 2026: Report.
The model is currently available to Plus, Pro, Enterprise, and Edu subscribers across ChatGPT Work and Codex, alongside standard API access. OpenAI confirmed that GPT-6.1 Terra retains safety protocols established under its Preparedness Framework, featuring improved resistance against prompt-injection attacks and better adherence to defined task boundaries.
(The above story first appeared on LatestLY on Oct 27, 2026 03:49 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).