OpenAI Launches GPT-6.1 Terra: Advanced Computer-Use Capabilities at a Fraction of the Cost

OpenAI Launches GPT-6.1 Terra: Advanced Computer-Use Capabilities at a Fraction of the Cost

On October 26, 2026, OpenAI introduced a major disruption to the AI enterprise market with the release of its new mid-tier model, GPT-6.1 Terra (gpt-6.1-terra). Built as part of the GPT-6.1 series, this model is engineered specifically for high-volume, reliable automation, focusing heavily on digital screen interaction (computer use) and visual reasoning.

Crucially, OpenAI has positioned Terra as an aggressively priced alternative for businesses, cutting API costs by approximately 50% compared to both GPT-6.1 Sol and the previous-generation GPT-5.6 Terra.

Core Capabilities and Features

1. Reliable Computer Use (Screen & App Automation) The standout feature of GPT-6.1 Terra is its "computer-use" capability, available as a hosted tool via the Responses API. The model can visually parse desktop interfaces, browsers, and applications, executing complex multi-step workflows like clicking buttons, switching between apps, and navigating digital environments. Early reports indicate it operates up to 3x faster than GPT-6.1 Sol and GPT-6 Astra in these specific environments.

2. Advanced Visual Reasoning Terra is highly adept at processing complex visual information, including intricate technical drawings, architectural blueprints, and dense UI screenshots. Thanks to its upgraded multimodal architecture, it has demonstrated exceptional performance on the MMMU-Pro benchmark, making it ideal for tasks requiring deep visual comprehension.

3. Enterprise Data Entry and Document Processing The model is optimized for extracting structured data from messy, unstructured visual sources like invoices, receipts, and complex spreadsheets. It can seamlessly ingest this data and automatically populate fields within accounting and ERP platforms like SAP or QuickBooks, while also possessing the natural language capabilities to quickly summarize email threads or sort customer support tickets.

Benchmark Performance

According to independent testing and OpenAI's internal benchmarks, GPT-6.1 Terra holds its own against significantly more expensive models in targeted areas:

• Terminal-Bench 4.0 (Coding/Command-line): At its maximum reasoning setting ("maximum effort"), the model scored a highly competitive 50.3%.

• HLE (Humanity's Last Exam): Achieved a 46.2% score, placing it firmly within the range of heavier flagship models.

• LCR (Long Context Reasoning): Scored a robust 82.7%, demonstrating strong adherence to instructions when processing massive data files.

• Agents' Last Exam (Computer use): Displayed professional-level proficiency in parsing technical patterns and executing tasks on medium reasoning settings.

API Pricing and Accessibility

OpenAI has priced Terra to encourage mass deployment across enterprise ecosystems:

• Base API Pricing: $1.00 per million input tokens and $5.00 per million output tokens.

• Prompt Caching: Highly discounted at just $0.05 per million cached input tokens.

• Batch Processing: Workloads run asynchronously via the Batch API receive an additional 50% discount.

• Context Window: The model supports a massive 1.05 million token context window, with a maximum output capacity of 128,000 tokens.

• Availability: Currently accessible to developers via the OpenAI API, and rolling out to users on ChatGPT Work, Codex, Plus, Pro, Enterprise, and Edu plans.