OpenAI Launches GPT-6.1 Terra at $1/M Input and $5/M Output

OpenAI introduces GPT-6.1 Terra, a fast model focused on visual reasoning and computational usage. Its pricing is set at $1 per million input tokens and $5 per million output tokens, which is nearly half the price of Sol or the previous generation, but ten times that of Luna. A "max" level example shows a cost of 11.3161 cents for 4 minutes and 11 seconds. Ted Sanders from OpenAI details the uses and positioning of the model.

Pricing, Comparisons, and an Example of Cost and Duration

GPT-6.1 Terra is offered at $1 per million input tokens and $5 per million output tokens. This pricing represents about half that of GPT-5.6 Terra and GPT-6.1 Sol, and it is approximately ten times higher than that of GPT-6 Luna. A comparative table presents the costs for each reasoning level of the GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna models. For a "max" reasoning level, the cost reaches 11.3161 cents, and the execution duration is 4 minutes and 11 seconds. Terra occupies a mid-range pricing position, similar to Anthropic's Sonnet model. Computational usage is cited as an area where a less expensive model can be advantageous, particularly for automated and repetitive tasks.

Announced Capabilities in Vision, Writing, and Computational Usage

OpenAI describes GPT-6.1 Terra as a fast and efficient model, combining detailed visual reasoning with precise computational usage. It is also presented as offering clearer and more natural writing, representing an advancement over GPT-5.6 Terra. Highlighted uses include drafting proposals, interpreting technical drawings, filling out forms, and video editing. The launch of GPT-6.1 Terra has been officially announced by OpenAI.

Recommendations from Ted Sanders for Developers and Subscribers

According to Ted Sanders from OpenAI, GPT-6.1 Terra likely surpasses Astra and Sol in technical communication, shows comparable performance in vision and computational tasks, while being significantly faster, but remains slightly behind on most other criteria. For API developers, he describes this model as an intermediate solution, offering lower costs than Sol and superior intelligence to Luna, particularly in the areas of vision, computational usage, and communication. For subscribers, he clarifies that this is not a revolutionary release, but the model can help overcome certain usage limits and provide clearer communication or faster computational usage.