OpenAI has introduced GPT-6.1 Terra, a model that combines detailed visual reasoning and precise computer use with clearer, more natural writing.
GPT-6.1 Terra
Compared with GPT-5.6 Terra, it improves performance on professional and software engineering tasks, reduces factual errors, and completes computer-based tasks faster. Its applications include drafting proposals, interpreting technical drawings, filling out forms, and editing videos.
Computer use and visual reasoning
OpenAI evaluated GPT-6.1 Terra on Agents’ Last Exam, a benchmark of professional tasks that includes converting architectural drawings into 3D models and reviewing an animated film shot by shot to create a time-coded log for a restoration team.
At medium reasoning effort, it scored 15 percentage points higher than GPT-5.6 Terra at approximately one-third of the cost. Its performance was also on par with GPT-6 Astra at one-tenth of the price per token.
At maximum reasoning effort, it recorded the following results compared with Claude Opus 5.5 and other models:
- Token efficiency: More than three times that of Claude Opus 5.5.
- Price per token: One-fourth that of Claude Opus 5.5.
- Cost per task: Approximately nine times lower than Claude Opus 5.5.
- Task completion speed: More than three times faster than GPT-6 Astra and GPT-6.1 Sol.
On MMMU-Pro, a visual reasoning benchmark, GPT-6.1 Terra outperformed GPT-5.6 Terra at every reasoning setting. It also nearly matched the scores of GPT-6 Astra and GPT-6.1 Sol at a substantially lower cost.
OSWorld 2.1 evaluates how well AI agents use desktop applications to complete multistep tasks, such as building a 3D model from a technical drawing. At medium reasoning effort, GPT-6.1 Terra surpassed GPT-5.6 Terra’s best score at a fraction of the cost per task.
Factual accuracy and software engineering
OpenAI said it has invested in reducing hallucinations and factual errors produced by its models. On its internal benchmark of error-inducing prompts, GPT-6.1 Terra recorded the following error rates:
- Low reasoning effort: 16.7%, compared with 53.6% for GPT-5.6 Terra, a reduction of nearly 37 percentage points.
- Maximum reasoning effort: 4.3%, a reduction of 18 percentage points, nearly matching the rates recorded by GPT-6 Astra and GPT-6.1 Sol.
The model was also evaluated on Terminal-Bench 4.0, which tests coding agents on command-line tasks, including fixing a webpage whose layout shifts as it loads. At maximum reasoning effort, GPT-6.1 Terra solved 50.3% of tasks, compared with GPT-5.6 Terra’s published score of 21.5%. OpenAI noted that GPT-6.1 Sol and GPT-6 Astra remain better choices for the most complex coding tasks.
Terminal-Bench-Science 0.1 evaluates how well agents use code for scientific research, including tracking cell divisions in a time-lapse video. At maximum reasoning effort, GPT-6.1 Terra surpassed GPT-6.1 Sol at a lower cost and came within five percentage points of Claude Opus 5.5 at less than one-fifth of the estimated cost.
Safety and alignment
GPT-6.1 Terra meets the Critical threshold in cybersecurity under OpenAI’s Preparedness Framework and maintains the safeguards developed for GPT-6 Astra. It also incorporates the alignment improvements of the GPT-6 series.
According to OpenAI’s evaluations, the model is more likely than GPT-5.6 Terra to stay within the scope of a task and honestly disclose limitations and potential blockers to users. In prompt-injection tests, it also demonstrated resistance to malicious instructions designed to divert it from the user’s task.
Pricing and availability
OpenAI said GPT-6.1 Terra’s lower API costs make it practical for developers to automate high-volume work, such as reading incoming invoices and entering their details into accounting software. In ChatGPT Work and Codex, the same usage allowance covers more tasks with GPT-6.1 Terra than with GPT-5.6 Terra, and substantially more than with GPT-6.1 Sol or GPT-6 Astra.
Availability
- ChatGPT Work and Codex: Available to Plus, Pro, Enterprise, and Edu users starting October 27, 2026.
- OpenAI API: Available to developers under the model identifier
gpt-6.1-terra.
Standard API pricing
- Input tokens: US$1 per million tokens.
- Cached input tokens: US$0.05 per million tokens.
- Output tokens: US$5 per million tokens.
These rates are 50% lower than those of GPT-5.6 Terra and GPT-6.1 Sol.