OpenAI Launches GPT-6.1 Terra with Advanced Computer Use Capabilities

Ender Öztürk

October 26, 2026

OpenAI officially introduced its latest artificial intelligence model, GPT-6.1 Terra, on October 26, 2026, marking a significant leap in visual reasoning and autonomous digital interaction. Designed to operate across various applications by understanding on-screen content, this model is positioned as a faster and more affordable alternative to its predecessors. GPT-6.1 Terra excels at handling routine workday tasks, such as filling out forms, interpreting technical drawings, and managing communications with more natural language. By reducing the barrier to entry for high-volume automation, OpenAI aims to empower developers and enterprises with a highly efficient tool for complex digital workflows.

According to the official announcement, GPT-6.1 Terra is designed to be the successor to the GPT-5.6 Terra model. It focuses on speed and efficiency, making it an ideal candidate for professional environments where time-sensitive tasks are common. The model can interpret visual data on a screen, allowing it to move between different applications to complete multi-step processes. Whether it is preparing a business proposal or editing a video, OpenAI suggests that this model is capable of handling a wide variety of computer-based assignments that previously required constant human intervention.

Operational Costs for Developers Significantly Decrease

One of the most striking aspects of the GPT-6.1 Terra release is its pricing structure. OpenAI has revealed that the standard API prices for this model are approximately 50 percent lower than those for GPT-5.6 Terra and GPT-6.1 Sol. Specifically, the model is priced at 1 USD per million input tokens, 0.05 USD per million cached input tokens, and 5 USD per million output tokens. This aggressive pricing strategy is intended to make high-volume automation practical for businesses of all sizes.

For instance, developers can now automate the process of reading incoming invoices and entering that data into accounting software without worrying about prohibitive costs. OpenAI claims that in the “Agents’ Last Exam” test, which measures performance on professional tasks, GPT-6.1 Terra scored 15 percentage points higher than GPT-5.6 Terra while operating at roughly one-third of the cost. Furthermore, when compared to competitors like Anthropic’s Claude Opus 5.5, OpenAI asserts that Terra offers a cost advantage that is nearly nine times lower per task at the highest reasoning levels.

Reasoning Benchmarks and Technical Accuracy Levels Rise

Beyond cost efficiency, GPT-6.1 Terra represents a major advancement in technical precision and software development. In internal tests involving prompts designed to trigger errors, the model showed a dramatic improvement. The error rate for low-reasoning tasks dropped from 53.6 percent in the previous generation to just 16.7 percent in GPT-6.1 Terra. At the highest reasoning levels, the error rate was recorded as low as 4.3 percent, showcasing the model’s reliability for critical operations.

Technical performance was also measured using the Terminal-Bench 4.0 test, which evaluates the ability to solve software issues through a command-line interface. At its highest reasoning level, the model successfully resolved 50.3 percent of the tasks presented. This is more than double the 21.5 percent success rate reported for GPT-5.6 Terra. Despite these gains, OpenAI notes that for the most complex coding projects, the GPT-6.1 Sol and GPT-6 Astra models remain the superior choices due to their specialized architectures.

New Security Protections and Global Access Features Arrive

Starting today, GPT-6.1 Terra is available to a wide range of users, including those on ChatGPT Plus, Pro, Enterprise, and Edu plans, as well as those using Codex. Developers can access the model via the OpenAI API under the identifier “gpt-6.1-terra.” The model features a massive 1,050,000-token context window, providing ample space for processing large documents or long-running digital interactions. Additionally, the model supports the computer use tool through the Responses API, allowing for deeper integration into third-party software.

Security remains a top priority for OpenAI with this release. The company stated that GPT-6.1 Terra has met the “Critical” threshold within its Preparedness Framework for cybersecurity. The model incorporates the same robust protections developed for GPT-6 Astra, ensuring that it can resist malicious instructions intended to divert the AI from its assigned tasks. These safety measures are designed to provide enterprises with the confidence needed to deploy autonomous agents in sensitive work environments.

How do you think autonomous computer use will change your daily productivity? Share your thoughts on the future of AI agents in the comments below.