Introducing GPT-6.1 Terra: A Faster, More Accurate AI Model Designed for Complex Automation

Launch of the New GPT-6.1 Terra Model

According to НВ — Техно: On October 27, OpenAI unveiled GPT-6.1 Terra, an advanced artificial intelligence model engineered to handle complex computational tasks, particularly in automating professional workflows. This latest iteration is significantly faster and more cost-effective than its predecessor, GPT-6 Astra. GPT-6.1 Terra demonstrates enhanced accuracy, fewer errors, and improved programming efficiency compared to earlier versions.

In benchmark tests, GPT-6.1 Terra matched the performance of GPT-6 Astra while offering token costs roughly ten times lower. Its primary applications include automating tasks such as:

• invoice processing;
• technical drawing analysis;
• interaction with desktop software.

During the Agents' Last Exam evaluation, GPT-6.1 Terra outperformed GPT-5.6 Terra by 15 percentage points, with operational costs approximately one-third of GPT-5.6 Terra’s expenses.

Key Features and Advantages of GPT-6.1 Terra

This model operates at speeds more than three times faster than both GPT-6 Astra and GPT-6.1 Sol. When operating at the highest reasoning level, GPT-6.1 Terra executes tasks at a cost nearly nine times lower than Claude Opus 5.5. In the OSWorld 2.1 test, GPT-6.1 Terra surpassed GPT-5.6 Terra at medium reasoning levels, and it outperformed the latter across all reasoning tiers in the MMMU-Pro test. Error rates at low reasoning levels were 16.7% for GPT-6.1 Terra, considerably better than the 53.6% observed with GPT-5.6 Terra. Under maximum reasoning conditions, GPT-6.1 Terra's error rate dropped to 4.3%, an 18 percentage point improvement over GPT-5.6 Terra.

In the Terminal-Bench 4.0 assessment, GPT-6.1 Terra successfully completed 50.3% of tasks at the highest reasoning level, compared to just 21.5% for GPT-5.6 Terra. For the most demanding programming challenges, OpenAI recommends using GPT-6.1 Sol or GPT-6 Astra.

Available since October 27, GPT-6.1 Terra can be accessed by ChatGPT Work and Codex users subscribed to Plus, Pro, Enterprise, or Edu plans. Developers can also integrate GPT-6.1 Terra via the OpenAI API using the identifier gpt-6.1-terra. API pricing is set at $1 per million input tokens, $0.05 per million cached input tokens, and $5 per million output tokens-rates that are 50% lower than those for GPT-5.6 Terra and GPT-6.1 Sol.

Security-wise, GPT-6.1 Terra incorporates the protective features of GPT-6 Astra, including reinforced defenses against adversarial prompts. OpenAI classifies this model as carrying a critical cybersecurity risk level. The reduced cost structure is expected to make advanced automation accessible to a broader range of enterprises and developers.

The release of GPT-6.1 Terra stands to reshape the AI landscape by combining cost reduction with enhanced performance, unlocking new possibilities for businesses aiming to integrate automation into their operations.

The advancements in AI technology continue to reshape various sectors, as seen in the recent improvements made by OpenAI. For instance, the latest model has achieved a remarkable 97.9% accuracy in migrating legacy code, demonstrating how these innovations can enhance programming efficiency and accuracy in complex tasks. Such developments highlight the growing importance of AI in optimizing workflows across industries.