GPT-6.1 Terra explained: price, benchmarks, builds.

GPT-6.1 Terra is OpenAI’s new fast, efficient model for computer use, visual reasoning and everyday writing. OpenAI pitches it as the model for the work that fills a normal day, from drafting proposals and reading technical drawings to filling out forms and editing videos, at a much lower price than its frontier models. It has also quickly become one of the most tested models on the Atlas, with creators racing it against Claude Opus 5.5 and GPT-6.1 Sol on the same prompts. Here is what OpenAI has confirmed about GPT-6.1 Terra, what it costs, how to try it, and what people are building with it.

What is GPT-6.1 Terra?

OpenAI describes GPT-6.1 Terra as a fast, efficient model that combines detailed visual reasoning and precise computer use with clearer, more natural writing. It replaces GPT-5.6 Terra, which OpenAI says it is a significant step up from. In OpenAI’s lineup, Terra is the cost-efficient option: for the most complex coding tasks, OpenAI still recommends GPT-6.1 Sol and GPT-6 Astra.

According to OpenAI’s API documentation, the model ID is gpt-6.1-terra. It accepts text and images and writes text (audio and video input are not supported), has a 1,050,000-token context window with up to 922,000 input tokens, can write up to 128,000 output tokens, and has a knowledge cutoff of 1 June 2026. Reasoning effort can be set to none, low, medium (the default), high, xhigh or max.

What’s new in GPT-6.1 Terra: computer use and visual reasoning

OpenAI’s headline claim is computer use at scale: an agent that looks at screens and documents and acts on them. OpenAI reports these results:

• Agents’ Last Exam: at medium reasoning effort, OpenAI says Terra scores 15 percentage points higher than GPT-5.6 Terra at about one-third of the cost, and performs on par with GPT-6 Astra at one-tenth of Astra’s price per token.
• Against Claude Opus 5.5: at max effort, OpenAI says Terra is over three times as token-efficient as Opus 5.5 and a quarter of its price per token, for roughly 9x lower cost per task on that benchmark.
• Speed: OpenAI says Terra completes tasks over three times faster than both GPT-6 Astra and GPT-6.1 Sol.
• MMMU-Pro (visual reasoning): OpenAI says Terra beats GPT-5.6 Terra at every reasoning setting and nearly matches GPT-6 Astra and GPT-6.1 Sol at much lower cost.
• OSWorld 2.1 (desktop apps): at medium effort, OpenAI says Terra surpasses GPT-5.6 Terra’s best score at a fraction of the cost per task.

Fewer factual errors and better coding

OpenAI says it invested heavily in reducing hallucinations this year. 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, and at maximum effort it fell by 18 percentage points to 4.3%, which OpenAI says nearly matches GPT-6 Astra and GPT-6.1 Sol.

For coding, OpenAI reports that GPT-6.1 Terra solves 50.3% of Terminal-Bench 4.0 tasks at maximum effort, more than double GPT-5.6 Terra’s published 21.5%. On Terminal-Bench-Science 0.1, OpenAI says Terra at max effort surpasses GPT-6.1 Sol at lower cost and comes within five points of Opus 5.5 at less than one-fifth of the estimated cost.

On safety, OpenAI says GPT-6.1 Terra meets the Critical threshold in cybersecurity under its Preparedness Framework and keeps the protections developed for GPT-6 Astra, and that in its tests Terra showed strong resistance to prompt injection. More detail is in OpenAI’s system card.

GPT-6.1 Terra pricing and availability

OpenAI says GPT-6.1 Terra is available to all Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and to developers in the OpenAI API, where it supports the Responses, Chat Completions and Batch endpoints. Standard API pricing, per OpenAI:

• Input: $1 per million tokens.
• Cached input: $0.05 per million tokens.
• Output: $5 per million tokens.
• Long prompts: requests with more than 272K input tokens are billed at 2x the input and cache rates and 1.5x the output rate for the whole request.
• Batch and Flex: 50% lower than Standard; Fast mode costs 2x Standard.

OpenAI says these standard API prices are 50% lower than GPT-5.6 Terra and GPT-6.1 Sol, and that in ChatGPT Work and Codex 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.

Who GPT-6.1 Terra is for

• Teams automating high-volume work: OpenAI’s own example is reading incoming invoices and entering their details into accounting software.
• Agent builders: the docs list computer use, hosted shell, apply patch, code interpreter, web search, file search and MCP among the tools Terra supports in the Responses API.
• ChatGPT Work and Codex users who want their plan’s allowance to stretch further on everyday tasks.
• Not the first pick for the hardest coding: OpenAI itself points to GPT-6.1 Sol and GPT-6 Astra for the most complex software work.

How to try GPT-6.1 Terra

1. In ChatGPT Work or Codex on a Plus, Pro, Enterprise or Edu plan, pick GPT-6.1 Terra as the model.
2. In the OpenAI API, call the model gpt-6.1-terra. Use the Responses API if you need tool calling, which OpenAI’s docs say requires it.
3. Start at the default medium reasoning effort, which is the setting behind several of OpenAI’s cost comparisons, and raise it for harder tasks.

What people are building with GPT-6.1 Terra

Creators on X have mostly put Terra through head-to-head tests, giving it the same prompt as a pricier model and posting the time and cost. These come from the creators’ own posts, not from OpenAI. You can browse them all on the GPT-6.1 Terra model page:

• Space train benchmark: GPT-6.1 Terra vs Sol: EnvolDev reports Terra finished 2.4x cheaper and about 20 minutes faster than GPT-6.1 Sol.
• Table tennis game: GPT-6.1 Terra vs Kimi K3.1: the creator reports Terra finished in 34 minutes against Kimi’s 48.
• Interactive steam locomotive cutaway: a rotatable 3D cutaway with live boiler pressure, piston mechanics and a throttle slider.
• Shaking oak tree: GPT-6.1 Terra vs Opus 5.5 and carvable jelly pumpkin: same-prompt comparisons against Claude Opus 5.5 with time and cost.
• Ink drop to jellyfish to galaxy: a single unbroken shot rendered by Terra and Opus 5.5.
• 3D supercar built by four frontier models and helicopter swashplate across four models: Terra lined up against several other models on one prompt.

To compare Terra with the models it is most often tested against, see the builds made with Claude Opus 5.5, GPT-6 Sol, GPT-6 Astra and Kimi K3.1 Pro.

GPT-6.1 Terra FAQ