OpenAI Launches GPT-6.1 Terra, a Cheaper AI Model Built for Computer Use
OpenAI's new model halves API prices compared with GPT-5.6 Terra while promising faster computer automation, better visual reasoning and fewer factual errors.
Published
Last updated / Fact-checked on
OpenAI has introduced GPT-6.1 Terra, a new artificial intelligence model designed to perform everyday computer tasks faster and at lower cost. Announced on Oct. 26, the model targets businesses and developers looking to automate workflows that involve software applications, documents and visual information.
The release marks a push to make AI agents more affordable for routine professional work, rather than focusing only on ever more powerful flagship models.
According to OpenAI, GPT-6.1 Terra improves on GPT-5.6 Terra in computer use, factual accuracy and software engineering, while cutting standard API prices roughly in half.
Designed to Work Across Computer Applications
GPT-6.1 Terra combines visual reasoning, computer interaction and natural-language writing to complete tasks that would otherwise require manual work. OpenAI's documentation highlights uses such as matching product photos to catalog entries, interpreting technical drawings, acting on application screens and drafting text.
These capabilities matter most to organizations building AI agents that work through software interfaces instead of relying solely on traditional integrations.
OpenAI evaluated the model on Agents' Last Exam, a benchmark of professional tasks that require software interaction and visual understanding, such as turning architectural drawings into a 3D model. At medium reasoning effort, OpenAI says GPT-6.1 Terra scored 15 percentage points higher than GPT-5.6 Terra at about one-third the cost.
The company also says Terra performed on par with GPT-6 Astra on the benchmark at one-tenth the price per token, and completed tasks more than three times faster than both GPT-6 Astra and GPT-6.1 Sol.
Lower Pricing Could Expand AI Automation
Cost is one of the release's biggest selling points. OpenAI's standard API pricing for GPT-6.1 Terra is:
| Usage category | Price per million tokens |
|---|---|
| Input | $1.00 |
| Cached input | $0.05 |
| Output | $5.00 |
OpenAI describes these rates as 50% lower than GPT-5.6 Terra and GPT-6.1 Sol. Its pricing page shows GPT-6.1 Sol at $2 input and $10 output, and GPT-5.6 Terra at $2 input and $12 output, so the cut is half or slightly more, depending on the comparison. Prompts longer than 272,000 tokens are billed at higher rates.
For developers running large numbers of automated tasks, lower token costs could make previously expensive workflows practical. Businesses processing invoices, maintaining product catalogs or automating administrative work may benefit most from the combination of lower prices and faster execution.
Improvements in Accuracy and Coding
OpenAI also reported a sharp drop in factual errors. On its internal set of prompts designed to provoke mistakes, GPT-6.1 Terra had a 16.7% error rate at low reasoning effort, compared with 53.6% for GPT-5.6 Terra. At maximum reasoning effort, the error rate fell to 4.3%.
Software engineering improved as well. On Terminal-Bench 4.0, which tests how coding agents handle command-line tasks, Terra solved 50.3% of tasks at maximum effort, compared with 21.5% for its predecessor. OpenAI still recommends GPT-6.1 Sol and GPT-6 Astra for the most demanding coding work.
The company says Terra meets the Critical cybersecurity threshold under its Preparedness Framework and includes safeguards developed for GPT-6 Astra.
All of these performance and safety figures come from OpenAI's own evaluations, so independent testing matters before organizations deploy the model in sensitive workflows.
Availability and What Comes Next
GPT-6.1 Terra is available to ChatGPT Plus, Pro, Enterprise and Edu users through ChatGPT Work and Codex. Developers can access it through OpenAI's API using the identifier gpt-6.1-terra.
The release comes amid growing competition among AI developers to deliver models that can carry out real-world tasks efficiently. More powerful models remain available for complex reasoning and research, but Terra's mix of speed, computer-use skills and lower pricing could make it attractive for high-volume business automation.
The larger question is whether those gains hold up across real workplace applications. For organizations evaluating AI agents, the release offers another way to balance capability, operating cost and speed.
Sources
- OpenAI, "Introducing GPT-6.1 Terra": https://openai.com/index/introducing-gpt-6-1-terra/
- OpenAI, GPT-6.1 Terra model documentation: https://developers.openai.com/api/docs/models/gpt-6.1-terra
- OpenAI, API pricing: https://developers.openai.com/api/docs/pricing
- BleepingComputer, "OpenAI launches GPT-6.1 Terra with 3x faster AI agents, 50% lower pricing": https://www.bleepingcomputer.com/news/artificial-intelligence/openai-launches-gpt-61-terra-with-3x-faster-ai-agents-50-percent-lower-pricing/
- The Verge, "GPT-6.1 Terra is very good at computer use, OpenAI says": https://www.theverge.com/ai-artificial-intelligence/1018854/gpt-6-1-terra-is-very-good-at-computer-use-openai-says