OpenAI launches GPT-6.1 Terra: it uses your computer, and the API costs half

What happened. On 26 October 2026 OpenAI released GPT-6.1 Terra. The company calls it a fast, efficient model that combines visual reasoning and precise computer use with clearer, more natural writing. It replaces GPT-5.6 Terra. 9to5Mac notes that OpenAI has four model classes, from cheapest to most capable: Luna, Terra, Sol and Astra.

What it does. OpenAI's examples: drafting proposals, reading technical drawings, filling out forms and editing videos. On the Agents' Last Exam benchmark at medium reasoning, it scores 15 percentage points higher than GPT-5.6 Terra at about one-third the cost. OpenAI says it finishes tasks over three times faster than GPT-6 Astra and GPT-6.1 Sol.

Fewer mistakes. On an internal set of error-inducing prompts, the error rate at low reasoning fell from 53.6% with GPT-5.6 Terra to 16.7%. At max effort it fell 18 points to 4.3%. On Terminal-Bench 4.0, a command-line coding test, it solves 50.3% of tasks versus 21.5% for the previous version. For the hardest coding work OpenAI still points to Sol and Astra.

Price and access. In the API the model is called gpt-6.1-terra. Standard pricing is $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens. A token is a small chunk of text, roughly part of a word. In ChatGPT Work and Codex it is available to Plus, Pro, Enterprise and Edu users.

Safety. OpenAI says the model meets the Critical threshold in cybersecurity under its Preparedness Framework and keeps the protections built for GPT-6 Astra. Details are in the model's system card.

What is not known. All benchmark results were published by OpenAI itself, with no independent checks yet. It is not known when free ChatGPT users will get it.

已确认的事实

• On 26 October 2026 OpenAI introduced the GPT-6.1 Terra model.^[1][2]
• OpenAI describes it as a fast, efficient model combining visual reasoning, precise computer use and more natural writing.^[1]
• Its standard API prices are 50% lower than those of GPT-5.6 Terra and GPT-6.1 Sol.^[1][2]
• API pricing is $1 per million input tokens, $0.05 per million cached input tokens and $5 per million output tokens.^[1]
• It is available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and in the API as gpt-6.1-terra.^[1][2]
• On Agents' Last Exam at medium reasoning it scores 15 percentage points higher than GPT-5.6 Terra at about one-third the cost.^[1]
• OpenAI says it completes tasks over three times faster than GPT-6 Astra and GPT-6.1 Sol.^[1]
• On an internal benchmark the error rate at low reasoning fell from 53.6% to 16.7%, and at max effort it fell 18 points to 4.3%.^[1][2]
• On Terminal-Bench 4.0 at max reasoning it solves 50.3% of tasks versus 21.5% for GPT-5.6 Terra.^[1]
• OpenAI has four main model classes: Luna, Terra, Sol and Astra.^[2]
• OpenAI says the model meets the Critical cybersecurity threshold under its Preparedness Framework and keeps the protections developed for GPT-6 Astra.^[1]

影响

• 已确认Automating office work through the API now costs half of what it did with the previous Terra.
• 已确认ChatGPT Work and Codex subscribers get more tasks out of the same allowance.
• 我们的解读The price cut may push rivals to lower prices on similar models.

仍未知

• Whether independent tests confirm OpenAI's numbers.
• Whether and when free ChatGPT users get the model.

影响到谁

Worldwide. ChatGPT and Codex users, developers and companies automating work with AI

来源

1. OpenAI: Introducing GPT-6.1 Terra 2026年10月26日 · 官方
2. 9to5Mac: OpenAI upgrades ChatGPT and Codex with new GPT-6.1 Terra model 2026年10月26日 · 媒体