OpenAI has introduced GPT-6.1 Terra, a new artificial intelligence model designed to handle everyday professional tasks with greater speed, improved accuracy and lower operating costs. The model focuses on writing, coding, visual understanding and computer-based workflows, offering an alternative for users who need capable AI assistance without paying the higher costs associated with more advanced models.
According to OpenAI’s announcement on October 26, 2026, GPT-6.1 Terra is designed to help users complete tasks such as preparing documents, interpreting technical drawings, filling out forms and editing videos. The company says the model represents a significant improvement over its predecessor, GPT-5.6 Terra.
One of the most notable features is its pricing. OpenAI says GPT-6.1 Terra’s standard API prices are 50% lower than those of GPT-5.6 Terra and GPT-6.1 Sol. The company also reports that the model can match GPT-6 Astra’s performance on a particular computer-use benchmark at approximately one-tenth of its price per token. <Cite refs={[“turn418775search1″,”turn418775search2”]}/>
Faster Performance and Advanced Computer Use
GPT-6.1 Terra is designed to do more than generate text or answer questions. It can interpret information displayed on computer screens and support workflows involving multiple applications, making it useful for automating repetitive professional tasks.
OpenAI highlights examples such as converting architectural drawings into 3D models and reviewing videos frame by frame to create detailed, time-coded records. These capabilities could benefit professionals working in design, administration, media production and technical analysis.
The model’s computer-use capabilities were evaluated using Agents’ Last Exam, a benchmark focused on complex professional tasks. OpenAI reports that GPT-6.1 Terra scored 15 percentage points higher than GPT-5.6 Terra at medium reasoning effort while costing approximately one-third as much per task.
The company also says GPT-6.1 Terra achieved performance comparable to GPT-6 Astra on this benchmark at around one-tenth the price per token. OpenAI further reports that Terra completed tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol in its comparisons. These figures reflect the company’s reported evaluations and may vary across real-world tasks. <Cite refs={[“turn418775search1″,”turn418775search9”]}/>
Improved Visual Understanding and Accuracy
Visual reasoning is another important area of development for GPT-6.1 Terra. The model is designed to interpret images, technical diagrams and other visual information, helping users extract useful details from materials that would otherwise require manual review.
OpenAI says Terra performed strongly on the MMMU-Pro benchmark, which evaluates multimodal understanding and reasoning. Its results approached those of more powerful models, including GPT-6 Astra, while operating at a lower cost.
The company has also reported improvements in factual accuracy compared with GPT-5.6 Terra. According to its internal evaluations, the error rate on difficult prompts fell from 53.6% to 16.7% at low reasoning effort. At maximum reasoning effort, the reported error rate decreased to 4.3%.
These results suggest progress in reducing incorrect responses, although benchmark performance does not guarantee that every answer will be accurate. Users should continue checking important information, particularly when working with financial, legal or technical material. <Cite refs={[“turn418775search1″,”turn418775search9”]}/>
Lower Costs Could Expand AI Automation
The lower pricing could make GPT-6.1 Terra particularly attractive to businesses and developers that process large volumes of information. Companies can use AI models for tasks such as extracting invoice details, organizing documents, updating records and handling repetitive computer-based work.
Reducing the cost of these operations could help organizations automate more tasks within the same budget. Faster processing may also allow teams to complete routine work more efficiently, leaving employees with additional time for activities that require human judgment.
However, the most suitable model depends on the complexity of the task. OpenAI continues to position GPT-6 Astra for demanding reasoning and professional workflows, while GPT-6.1 Terra is aimed at fast, efficient everyday work. GPT-6.1 Sol provides another option for users seeking a balance between performance and cost. <Cite refs={[“turn418775search2″,”turn418775search4”]}/>
Availability and API Pricing
OpenAI has made GPT-6.1 Terra available to Plus, Pro, Enterprise and Edu users through ChatGPT Work and Codex. Developers can also access the model through the OpenAI API using the model identifier gpt-6.1-terra.
The published standard API pricing is $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens. Actual costs depend on the amount of information processed, the responses generated and the use of cached inputs. <Cite refs={[“turn418775search1″,”turn418775search2”]}/>
Overall, GPT-6.1 Terra represents OpenAI’s effort to make advanced AI capabilities more economical for routine professional tasks. Its reported improvements in computer use, visual reasoning, speed and pricing could make it a practical choice for businesses and developers seeking to expand AI-powered automation.
Although lower costs are a major advantage, users should evaluate the model against their own workloads before switching. The balance between speed, accuracy, capability and price will ultimately determine how useful GPT-6.1 Terra is for different applications.