TL;DR: GPT-6.1 Terra is OpenAI’s new mid-range model, released on October 26, 2026, and built for computer use: reading screens, clicking through apps and filling in forms. Standard API pricing is $1 per million input tokens and $5 per million output tokens, half the rate of GPT-5.6 Terra and GPT-6.1 Sol. OpenAI says it matches GPT-6 Astra on its Agents’ Last Exam benchmark at a tenth of the price per token. It is live now in ChatGPT Work and Codex for Plus, Pro, Enterprise and Edu users, and in the API as gpt-6.1-terra.
OpenAI shipped GPT-6.1 Terra on Monday, the same day Anthropic filed its IPO paperwork, so it was easy to miss. It is worth a closer look. In its launch post, OpenAI describes Terra as “a fast, efficient model that combines detailed visual reasoning and precise computer use alongside clearer, more natural writing.” Put simply, it is the model OpenAI wants you to use when an AI agent has to do repetitive work on a computer and the bill matters more than squeezing out the last few points of intelligence.
Below is what Terra is, what it costs, how it compares with the rest of OpenAI’s lineup, and which of the headline claims are OpenAI’s own numbers that still need independent testing.
What GPT-6.1 Terra is for
OpenAI’s current models come in tiers. If you need a refresher on the naming, our guide to GPT-6 Sol vs Luna explains how the families differ. Terra sits between the small, cheap Luna models and the more capable Sol and Astra models. OpenAI calls GPT-6.1 Terra “a significant step up from GPT-5.6 Terra” and lists the kind of jobs it has in mind:
- Drafting proposals and other routine business writing
- Interpreting technical drawings
- Filling out forms and acting on application screens
- Editing videos and reviewing footage shot by shot
- High-volume automation, such as reading incoming invoices and typing their details into accounting software
The developer model page adds examples such as matching product photos to catalog entries. In the Responses API, Terra supports tools including web search, file search, code interpreter, MCP and computer_use, the tool that lets a model look at a screen and operate it.
GPT-6.1 Terra pricing
Price is the main selling point. According to OpenAI’s API pricing page, these are the standard short-context rates per million tokens (if tokens are new to you, see what tokens are in AI):
| Model | Input | Cached input | Output |
|---|---|---|---|
| GPT-6 Astra | $10.00 | $1.00 | $50.00 |
| GPT-6.1 Sol | $2.00 | $0.10 | $10.00 |
| GPT-6.1 Terra | $1.00 | $0.05 | $5.00 |
| GPT-5.6 Terra | $2.00 | $0.20 | $12.00 |
| GPT-6 Luna | $0.10 | $0.01 | $0.50 |
A few details from the model page are easy to overlook:
- Prompts longer than 272K input tokens are billed at 2x the input and cache rates and 1.5x output for the whole request.
- Batch and Flex processing are 50% cheaper than Standard; Fast mode costs 2x Standard.
- Regional processing adds a 10% premium where it is offered.
- Cache writes cost $1.25 per million tokens for 30-minute retention and $2.50 for 12-hour retention.
Developer Simon Willison summed up the positioning in his write-up: Terra is “almost half the price of GPT-5.6 Terra and GPT-6.1 Sol, and around 10x the price of GPT-6 Luna.”
What OpenAI claims about performance
All of the benchmark figures below come from OpenAI’s own announcement. They have not yet been reproduced by outside testers, so treat them as the company’s claims for now.
Computer use
On Agents’ Last Exam, a set of professional tasks such as turning architectural drawings into a 3D model, OpenAI says GPT-6.1 Terra at medium reasoning effort scores 15 percentage points higher than GPT-5.6 Terra “at about one-third the cost,” and performs on par with GPT-6 Astra “at only one-tenth its price per token.” At maximum effort, OpenAI says Terra is over three times as token-efficient as Anthropic’s Claude Opus 5.5 at a quarter of its price per token, which works out to “a ~9x lower cost per task.” The company also says Terra finishes tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol.
Factuality
OpenAI says that 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 it reports 4.3%. Fewer made-up answers matter a lot for agents that act without a human checking each step; our explainer on AI hallucination covers why models get things wrong in the first place.
Coding
On Terminal-Bench 4.0, which tests coding agents working from the command line, OpenAI reports 50.3% of tasks solved at maximum effort, more than double GPT-5.6 Terra’s 21.5%. OpenAI is clear about the limits here: “For the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra remain the better choice.”
Safety notes
OpenAI says GPT-6.1 Terra “meets the Critical threshold in cybersecurity under our Preparedness Framework” and keeps the protections it built for GPT-6 Astra. It also says Terra is more likely than its predecessor to stay within scope, disclose its limits, and resist prompt-injection attacks. Prompt injection, where text on a web page or document tries to hijack an agent, is the main practical risk for any model that browses and clicks on your behalf, so it is worth testing before you hand Terra access to real accounts. We covered a related shift when ChatGPT gained access to Chrome tabs.
Where you can use it
- ChatGPT Work and Codex: available now to Plus, Pro, Enterprise and Edu users. OpenAI says the same usage allowance covers more tasks with Terra than with GPT-5.6 Terra, and “substantially more” than with GPT-6.1 Sol or GPT-6 Astra.
- OpenAI API: model ID
gpt-6.1-terra. - Third parties: GPT-6.1 Terra and a Terra Pro variant were listed on OpenRouter on launch day, according to the WhatIsTheLatest.ai digest.
Should you switch to GPT-6.1 Terra?
It depends on the job. Some quick rules of thumb based on how OpenAI positions the model:
- Repetitive screen and form work: Terra is the obvious first model to try. This is exactly the work OpenAI built it for, and the price gap with Sol and Astra is large at volume.
- Simple text tasks at huge scale: Luna is still about 10x cheaper per token. If the work doesn’t need vision or computer use, test Luna first.
- Hard coding problems: OpenAI itself recommends Sol or Astra.
- Cost per task, not cost per token: a model that uses fewer tokens to finish can be cheaper even at a higher list price. Measure total spend on your own workload before you commit.
Terra is OpenAI’s bet that most businesses will pay for “good enough and quick” agents long before they pay for the very best model on every task. Whether its benchmark claims hold up will become clear once independent testers and early customers have had a few weeks with it.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
