OpenAI has introduced GPT-6.1 Terra, a new artificial intelligence model designed to help people and businesses complete computer-based tasks more quickly and efficiently. Announced on October 26, the model combines visual reasoning, computer-use capabilities, coding improvements, and natural-language writing in a more cost-effective package.
Thank you for reading this post, don't forget to subscribe!The launch highlights a growing shift in the AI industry: companies are moving beyond chatbots that simply answer questions toward AI systems that can help carry out practical, multi-step tasks.
According to OpenAI, GPT-6.1 Terra is built for everyday professional activities, including preparing documents, interpreting technical drawings, completing online forms, editing videos, and processing business records. Developers can also use the model to build applications that automate repetitive work.
The model is available through ChatGPT Work and Codex for eligible subscribers, as well as through OpenAI’s API for developers.
What Makes GPT-6.1 Terra Different?
Traditional AI assistants primarily respond to prompts with text, images, or code. Models with computer-use capabilities are designed to go further by interpreting what appears on a screen and taking actions across digital applications.
GPT-6.1 Terra is intended to combine these capabilities with improved reasoning and clearer writing.
For example, a business could use an AI-powered workflow to read incoming invoices, extract relevant information, and enter the details into accounting software. A creative professional could use the model to review video footage and prepare a time-coded log. Engineers and designers could also use visual reasoning to work with technical drawings and other complex visual material.
These examples illustrate the potential of AI-powered automation. However, the results still depend on the task, the software being used, and the safeguards applied to the workflow.
OpenAI says the new model is substantially faster and more economical than some of its larger models for supported computer-based tasks.
OpenAI Reports Major Improvements in Speed and Cost
One of the most important aspects of GPT-6.1 Terra is its reported performance relative to its cost.
OpenAI says the model scored 15 percentage points higher than GPT-5.6 Terra on the Agents’ Last Exam benchmark at medium reasoning effort, while costing approximately one-third as much for the measured work. The company also reports that GPT-6.1 Terra performed on par with GPT-6 Astra on this benchmark at roughly one-tenth of its price per token.
OpenAI further states that GPT-6.1 Terra completed tasks on the benchmark more than three times faster than GPT-6 Astra and GPT-6.1 Sol.
These figures come from OpenAI’s evaluations and should not be interpreted as guarantees that every real-world task will be completed at the same speed or cost.
Nevertheless, the results suggest that smaller, more efficient AI models could become increasingly useful for companies that need to automate large numbers of routine tasks without paying the cost of using their most powerful models for everything.
Fewer Factual Errors and Better Coding Performance
AI systems can sometimes produce incorrect information or make mistakes when following complex instructions. Improving reliability is therefore an important challenge for AI developers.
OpenAI reports that GPT-6.1 Terra reduced its error rate on an internal benchmark involving difficult, error-inducing prompts from 53.6% with GPT-5.6 Terra to 16.7% at low reasoning effort.
At maximum reasoning effort, OpenAI reports an error rate of 4.3% for GPT-6.1 Terra, an improvement of 18 percentage points compared with its predecessor.
The company has also reported substantial progress in software engineering. On Terminal-Bench 4.0, a benchmark that evaluates coding agents on command-line tasks, GPT-6.1 Terra achieved a score of 50.3% at maximum reasoning effort, compared with 21.5% for GPT-5.6 Terra’s published score.
These results could make the model useful for developers working on debugging, software maintenance, and other programming tasks. OpenAI nevertheless recommends its more capable models, including GPT-6.1 Sol and GPT-6 Astra, for some of the most demanding coding and research work.
GPT-6.1 Terra Pricing: What Developers Need to Know
Cost is a central part of the new model’s positioning. OpenAI lists the following standard API prices:
- Input tokens: $1 per million tokens.
- Cached input tokens: $0.05 per million tokens.
- Output tokens: $5 per million tokens.
OpenAI says the standard API prices are 50% lower than those of GPT-5.6 Terra and GPT-6.1 Sol.
For companies processing large volumes of documents, handling customer requests, or automating administrative work, lower model costs could make it easier to deploy AI across more parts of their operations.
However, the total cost of an AI workflow may also include software integrations, data storage, human review, and other services. Businesses should test their own workloads before estimating potential savings.
How GPT-6.1 Terra Could Affect Businesses
The model’s computer-use and visual-reasoning capabilities point toward a future in which AI tools help complete entire workflows rather than simply provide advice.
Small businesses could explore automation for document processing, reporting, and data entry. Software developers could use AI to assist with debugging and routine coding. Media companies could investigate AI-assisted video analysis, research, and content preparation.
These applications could save time, but they also introduce practical considerations. Businesses must protect confidential information, verify important outputs, and decide which actions require human approval.
AI agents should be given only the access they need. Financial transactions, sensitive records, account changes, and other consequential actions may require additional checks before they are completed.
The strongest results are likely to come from combining automation with human oversight rather than treating AI output as automatically correct.
Availability: Where Can Users Access GPT-6.1 Terra?
OpenAI says GPT-6.1 Terra is available to eligible ChatGPT Plus, Pro, Enterprise, and Edu users through ChatGPT Work and Codex. Developers can access the model through the OpenAI API using the model identifier gpt-6.1-terra.
Availability and usage limits may depend on the product, subscription, and account configuration.
The release gives developers and businesses another option when choosing between advanced reasoning, speed, and operating costs. Users who need the highest level of performance for particularly difficult tasks may still prefer a larger model, while Terra is positioned for efficient work across a broad range of computer-based activities.
The Bigger Picture: AI Is Moving From Answers to Actions
GPT-6.1 Terra reflects a wider direction in artificial intelligence: the development of systems that can interpret digital environments, follow multi-step instructions, and assist with work across applications.
This approach could change how people use AI at work. Instead of manually copying information between tools, preparing every report from scratch, or completing repetitive administrative tasks, users may increasingly delegate parts of these processes to AI agents.
Yet greater autonomy also increases the importance of reliability, security, and clear limits on what an AI system can do.
For businesses considering adoption, the key question is not simply whether a model is powerful. It is whether the model can complete a specific task accurately, securely, and at a cost that makes sense.
GPT-6.1 Terra’s reported improvements make it a notable development in the race to build more practical and affordable AI systems. Its real-world impact will depend on how well those capabilities translate into everyday work.
