OpenAI launches GPT-6.1 Terra at 50% lower API cost
OpenAI launched gpt-6.1 terra on October 26, 2026, cutting API costs by 50% to target high-volume business tasks. Available in ChatGPT Work, Codex, and via API for developers, the model handles invoice entry, form filing, and video editing. For businesses, the update lowers costs by 50% for high-volume enterprise automation tasks compared to previous models.
How gpt-6.1 terra enterprise automation handles daily workflows
The rollout on October 26, 2026, expands access across several tiers of OpenAI tools. The model was made available on Monday to Plus, Pro, Enterprise, and Edu subscription tiers inside ChatGPT Work and Codex. Developers can also access the system through the standard OpenAI API. The deployment provides a step up in capability compared to GPT-5.6 Terra while offering lower operational expenses.
How gpt-6.1 terra api pricing compares to previous models
Examining standard API costs reveals a direct 50% price reduction for integration compared to previous offerings. The API rates are half those of both GPT-5.6 Terra and GPT-6.1 Sol. OpenAI stated that the release advances the cost-intelligence frontier across its model families, providing a new balance of speed, capability, and expense for automated workflows.
Evaluating gpt-6.1 terra vs gpt-6.1 sol for enterprise tasks
Comparing the new system to Sol highlights key differences in cost structure and intended workload focus. While both belong to the latest generation of OpenAI offerings, the new system halves the standard API cost of the Sol model. It provides stronger performance than GPT-5.6 Terra while maintaining lower expenses, allowing businesses to select the right capability level for high-volume enterprise tasks without overpaying for compute.
What it changes for online stores and enterprises
For businesses, the model reduces the direct expense of high-volume administrative workflows by 50%. Companies running automated pipelines for reading invoices, entering structured data into accounting software, filling out digital forms, drafting client proposals, interpreting technical drawings, and editing videos can process twice the volume for the same API cost.
Revenue growth and agentic workspace capabilities
The release follows significant financial growth driven by corporate adoption. Bloomberg reported on October 9 that annualized revenue for OpenAI is expected to reach at least $70 billion by the end of 2026, rising from $50 billion at the end of September. This revenue growth is largely driven by enterprise adoption.
Earlier in the year, OpenAI began rolling out ChatGPT Work on July 9. Designed as an agent for complex multi-step projects, ChatGPT Work interacts across user applications and files, spending hours on single tasks to generate finished materials like sheets, slides, documents, and web applications.
Strategic model pricing across enterprise workloads
The latest launch builds on price adjustments introduced earlier in the year to improve performance per dollar. In July, OpenAI reduced the API cost of GPT-5.6 Luna by 80% and lowered GPT-5.6 Terra by 20%. During the same update, the company increased the execution speed of GPT-5.6 Sol in the API without raising its price, establishing a tiered pricing model that lets organizations apply targeted intelligence at every operational stage.
Questions and answers
What is the new model used for in enterprise automation?
The model automates high-volume business operations, including reading invoices and entering information into accounting software. It also handles computer-based and communication tasks such as drafting proposals, interpreting technical drawings, filling out forms, and editing videos.
How much does the model cost compared to GPT-6.1 Sol?
Standard API prices for the new model are 50% lower than those of GPT-6.1 Sol and GPT-5.6 Terra, cutting API expenses in half for high-volume workloads.
Is the new model available in ChatGPT Work and Codex?
Yes. On October 26, 2026, OpenAI made the model available to Plus, Pro, Enterprise, and Edu users within ChatGPT Work and Codex, as well as to developers via the OpenAI API.
MOZGII view
How we would build it
If we were building an enterprise document and workflow system today, we would route high-volume tasks like reading invoices and filling out forms to gpt-6.1 terra. This would keep operational expenses low across standard accounting pipelines. We would send multi-step tasks, such as interpreting technical drawings or drafting client proposals, to gpt-6.1 sol or ChatGPT Work. Human reviewers would verify all parsed financial data before entry into accounting software.
The system would break on poorly scanned documents, non-standard layout changes, or corrupted video files. We would track the human override rate as the main metric to ensure low model accuracy does not create manual bottlenecks. Next, we would build automated routing rules that escalate failed extractions to higher-tier models before alerting human staff.
Sources
Where the facts in this story come from.