OpenAI launches GPT-6.1 Terra, a cheaper model for computer use

OpenAI launched GPT-6.1 Terra on October 26, 2026, a model it describes as "a fast, efficient model that combines detailed visual reasoning and precise computer use alongside clearer, more natural writing." It is in ChatGPT Work, Codex and the OpenAI API, where the model page lists the ID gpt-6.1-terra at $1 per 1M input tokens and $5 per 1M output tokens, with a 1,050,000-token context window. The Verge reported the launch under the headline "GPT-6.1 Terra is very good at computer use, OpenAI says."

#What changed?

OpenAI pitches Terra at everyday computer work. It says the model "excels at the communication and computer-based tasks that fill the workday, from drafting proposals and interpreting technical drawings to filling out forms and editing videos."

Details from the API model page:

The page says: "Tool calling requires the Responses API." It also notes that "Batch and Flex prices are 50% lower than Standard." The API's Free tier is not supported.

#How cheap is it?

OpenAI says: "Its standard API prices are 50% lower than those of GPT-5.6 Terra and GPT-6.1 Sol, making it practical for developers to automate high-volume work, like reading incoming invoices and entering their details into accounting software."

OpenAI's pricing table lists these standard prices per 1M tokens:

Our analysis: against GPT-6.1 Sol, Terra is half the price on both input and output. Against GPT-5.6 Terra, input is half, and output is $5.00 instead of $12.00, about 58% lower.

In ChatGPT Work and Codex, OpenAI says "the same usage allowance covers more tasks than with GPT‑5.6 Terra and substantially more than with GPT‑6.1 Sol or GPT‑6 Astra."

#What do OpenAI's benchmarks show?

All scores below are OpenAI's own tests, from the announcement text:

• Agents' Last Exam, which OpenAI calls "a benchmark of professional tasks": OpenAI says "At medium reasoning effort, GPT-6.1 Terra scores 15 percentage points higher than GPT-5.6 Terra at about one-third the cost." It adds: "It even performs on par with GPT-6 Astra at only one-tenth its price per token."
• Claude Opus 5.5: OpenAI claims that "At max effort, Terra is over three times as token efficient as Claude Opus 5.5, and one-fourth its price per token, resulting in a ~9x lower cost per task."
• Terminal-Bench 4.0: "At maximum reasoning effort, GPT‑6.1 Terra solves 50.3% of tasks, more than double GPT‑5.6 Terra’s published score of 21.5%."
• Error rate: on what OpenAI calls its "internal benchmark of error-inducing prompts," the error rate at low effort "fell from 53.6% with GPT-5.6 Terra to 16.7% with GPT-6.1 Terra."
• OSWorld 2.1 and MMMU-Pro: OpenAI says Terra beats GPT-5.6 Terra on both but gives no scores in the text.

OpenAI also says: "For the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra remain the better choice." No independent benchmark results were in our research.

#Who can use it and when?

OpenAI's sources describe the rollout differently:

• The announcement: "Starting today, GPT-6.1 Terra is available to all Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex." It adds: "Developers can also access it through the OpenAI API as gpt-6.1-terra."
• @OpenAI, at 23:36 IST on October 26, 2026: "GPT-6.1 Terra is rolling out now and is available to all Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex."
• @OpenAIDevs, at 23:37 IST: "GPT-6.1 Terra is also coming to the API and Codex."
• The API changelog, dated October 26: "Released" GPT-6.1 Terra for the Responses and Chat Completions APIs.
• The help-center release notes: "GPT-6.1 Terra is rolling out in ChatGPT Work and Codex." They add: "As it reaches your account, you can use Terra in the Work and Codex experiences available with your plan, region, and workspace settings."

The plan lists also differ. The announcement and @OpenAI name Plus, Pro, Enterprise and Edu. Four other OpenAI pages, the ChatGPT and Codex changelog, the What's new page, the Work and Codex pricing page and the help-center release notes, also list Business: "Plus, Pro, Business, Enterprise, and Edu". No source lists Free or Go.

Limits stated by OpenAI:

• Not in Chat. OpenAI's changelog says: "The model isn’t available in Chat and does not support UAE inference residency."
• Enterprise. "Enterprise access depends on workspace availability and administrator controls."
• Usage limits. The release notes say: "Your plan’s overall usage limits are unchanged."
• Codex CLI. "GPT-6.1 Terra requires Codex CLI 0.159.0 or later."

#How to get it

1. API. Call gpt-6.1-terra through the Responses API for tool calling. OpenAI's GPT-6 guide says: "Start with medium and increase it if the improvement in quality is worth the added time and tokens."
2. Migrating. "When migrating from GPT-5.6 Terra, compare both models on representative images and workflows using the same reasoning effort."
3. ChatGPT. "Choose GPT-6.1 Terra from the model list in Advanced in the ChatGPT desktop app or ChatGPT Work on the web." The changelog adds: "Work also supports Terra on mobile."
4. Codex CLI. Update to 0.159.0 or later, then run codex -m gpt-6.1-terra.

#What about safety?

OpenAI says: "GPT‑6.1 Terra meets the Critical threshold in cybersecurity under our Preparedness Framework, and maintains the protections developed for GPT‑6 Astra." Its system card addendum says: "Under our Preparedness Framework, we are treating GPT-6.1 Terra as Critical in cybersecurity and High for Biological and Chemical capability." Both are OpenAI's own ratings.

OpenAI also says: "In our prompt-injection tests, GPT-6.1 Terra also showed strong resistance to malicious instructions designed to divert it from the user’s task."

#How does it compare?

If you let Codex act on your machine, see our guide to controlling an AI coding agent's network access on a Mac. For OpenAI's handling of agent risks, see our coverage of OpenAI agents weakening misalignment monitoring and OpenAI training frontier models offline.

#What is still unclear?

• Rollout timing. @OpenAIDevs says "coming"; other OpenAI pages say released or rolling out. OpenAI had not reconciled them in the sources we checked.
• Business plan. OpenAI has not explained why the announcement leaves it out.
• Regions. OpenAI gives no country list, only the UAE inference residency exclusion.
• Independent tests. Every benchmark number is OpenAI's.