Sam Altman Announces OpenAI’s Budget-Friendly GPT-6.1 Terra, Touts ‘New Era’ of Computer Use

On Monday, OpenAI CEO Sam Altman announced GPT-6.1 Terra, a faster and cheaper AI model built to carry out tasks on computers.

Altman Highlights GPT-6.1 Terra’s Ability to Use Computers

Altman announced the model on X, writing, "6.1 Terra is here! It is good at a lot of things, but we are especially excited for this new era of computer use."

OpenAI introduced GPT-6.1 Terra as a more affordable option in its GPT-6 family of models.

The model is built to work with what appears on computer screens and handle tasks such as filling out forms, processing invoices, editing videos and understanding technical drawings.

OpenAI Says the Model Is Faster and Makes Fewer Errors

OpenAI said GPT-6.1 Terra can finish tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol.

The company also said its own tests found that the model made errors on 16.7% of difficult prompts when using a low level of reasoning effort. By comparison, GPT-5.6 Terra had an error rate of 53.6% in the same type of testing.

These results come from OpenAI's own tests and have not been checked by independent researchers.

The company said the model is also better at staying within the limits of a task, spotting obstacles and resisting prompt-injection attacks. These attacks try to trick an AI into following harmful or unwanted instructions.

However, OpenAI warned that Terra is not the best choice for every task. The company said, "For the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra remain the better choice."

GPT-6.1 Terra API Pricing Starts at $1 Per Million Input Tokens

Developers who use GPT-6.1 Terra through the API will pay $1 for every million input tokens and $5 for every million output tokens. These prices are half the listed rates for GPT-6.1 Sol.

GPT-6.1 Terra is now available to Plus, Pro, Enterprise and Edu subscribers through ChatGPT Work and Codex.

OpenAI recommends testing Terra with existing tasks before moving applications to the new model. Medium noted that requests with more than 272,000 input tokens are charged at higher rates for the entire request.

Disclaimer: This content was partly created with the help of AI tools and was reviewed and published by Benzinga editors.

Photo courtesy of Antonello Marangi via Shutterstock.com.