[{"data":1,"prerenderedAt":14},["ShallowReactive",2],{"artikel-en-berita-openai-gpt-6-1-terra-komputer-27-oktober-2026":3},{"slug":4,"lang":5,"title":6,"description":7,"date":8,"updated":9,"topic":10,"product":9,"cover":11,"readMin":12,"html":13},"berita-openai-gpt-6-1-terra-komputer-27-oktober-2026","en","OpenAI launches GPT-6.1 Terra, a fast model for computer tasks","OpenAI released GPT-6.1 Terra on October 26, 2026, a cheaper and faster model for computer use, writing and visual reasoning.","2026-10-27",null,"berita","\u002Fimages\u002Fartikel\u002Fberita-openai-gpt-6-1-terra-komputer-27-oktober-2026.webp",3,"\u003Cp>OpenAI launched GPT-6.1 Terra on October 26, 2026. The company describes it as a fast, efficient model that combines detailed visual reasoning, precise computer use, and more natural writing. It is now rolling out to users of ChatGPT Work and Codex.\u003C\u002Fp>\n\u003Ch2>1. What was launched\u003C\u002Fh2>\n\u003Cp>GPT-6.1 Terra is the successor to GPT-5.6 Terra in OpenAI's model family. The company built it for everyday office tasks: drafting proposals, interpreting technical drawings, filling out forms, and editing videos.\u003C\u002Fp>\n\u003Cp>It is available in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu plans. Developers can access it through the OpenAI API as \u003Ccode>gpt-6.1-terra\u003C\u002Fcode>.\u003C\u002Fp>\n\u003Cp>One note: Terra is only available in Work and Codex, not the regular Chat tab. This is confirmed in the ChatGPT Business release notes.\u003C\u002Fp>\n\u003Ch2>2. Prices drop by half\u003C\u002Fh2>\n\u003Cp>OpenAI says standard API pricing for GPT-6.1 Terra is 50% lower than GPT-5.6 Terra and GPT-6.1 Sol. The fixed rates are $1 per million input tokens, $0.05 per million cached input tokens, and $5 per million output tokens.\u003C\u002Fp>\n\u003Cp>In ChatGPT Work and Codex, the same usage allowance covers more tasks than with GPT-5.6 Terra. For high-volume work like reading incoming invoices and entering their details into accounting software, this makes automation more practical.\u003C\u002Fp>\n\u003Ch2>3. Computer-use performance\u003C\u002Fh2>\n\u003Cp>On the Agents' Last Exam benchmark, which tests professional tasks such as turning architectural drawings into a 3D model, GPT-6.1 Terra at medium reasoning effort scored 15 percentage points higher than GPT-5.6 Terra at about one-third the cost.\u003C\u002Fp>\n\u003Cp>OpenAI also says Terra performs on par with GPT-6 Astra at one-tenth its price per token, and completes tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol. At maximum effort, OpenAI claims Terra is over three times as token efficient as Claude Opus 5.5, for roughly 9x lower cost per task. These figures come from OpenAI itself, so treat them as vendor claims, not independent tests.\u003C\u002Fp>\n\u003Ch2>4. Fewer factual errors and stronger coding\u003C\u002Fh2>\n\u003Cp>OpenAI says on its internal benchmark of error-inducing prompts, the error rate at low reasoning effort fell from 53.6% with GPT-5.6 Terra to 16.7% with GPT-6.1 Terra. At maximum effort, the error rate fell 18 percentage points to 4.3%.\u003C\u002Fp>\n\u003Cp>For coding, Terminal-Bench 4.0 shows GPT-6.1 Terra solving 50.3% of tasks at maximum reasoning effort, versus GPT-5.6 Terra's published score of 21.5%. OpenAI itself advises that for the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra remain the better choice.\u003C\u002Fp>\n\u003Ch2>5. Safety and availability\u003C\u002Fh2>\n\u003Cp>GPT-6.1 Terra meets the Critical threshold in cybersecurity under OpenAI's Preparedness Framework and keeps the protections built for GPT-6 Astra. In prompt-injection tests, the model showed strong resistance to malicious instructions.\u003C\u002Fp>\n\u003Cp>The release notes list one limitation: Terra does not support UAE inference residency, a restriction based on where inference runs, not your location. Users can select GPT-6.1 Terra from the model list in Work or Codex, and overall plan usage limits are unchanged.\u003C\u002Fp>\n\u003Ch2>What it means for you\u003C\u002Fh2>\n\u003Cp>Cheaper models mean automation that used to be too expensive now makes sense. If you use ChatGPT for work, Terra may already be in your account.\u003C\u002Fp>\n\u003Col>\n\u003Cli>\u003Cstrong>Check the model list\u003C\u002Fstrong> in ChatGPT Work or Codex. If Terra is available, try it for visual and computer-based tasks.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>For developers:\u003C\u002Fstrong> the $1 per million input token price makes high-volume automation, like invoice processing, cheaper than before.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Skip it for the hardest coding.\u003C\u002Fstrong> OpenAI itself says Sol and Astra are still better for those tasks.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Watch your usage allowance:\u003C\u002Fstrong> a more efficient model means your plan covers more work under the same limits.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Access via API\u003C\u002Fstrong> as \u003Ccode>gpt-6.1-terra\u003C\u002Fcode> if you build your own products.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fopenai.com\u002Findex\u002Fintroducing-gpt-6-1-terra\u002F\" target=\"_blank\" rel=\"noopener\">Introducing GPT-6.1 Terra\u003C\u002Fa> (OpenAI)\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F6825453-chatgpt-release-notes\" target=\"_blank\" rel=\"noopener\">ChatGPT release notes\u003C\u002Fa> (OpenAI Help Center)\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F11391654-chatgpt-business-release-notes\" target=\"_blank\" rel=\"noopener\">ChatGPT Business release notes\u003C\u002Fa> (OpenAI Help Center)\u003C\u002Fli>\n\u003C\u002Ful>\n",1793543301932]