# GPT-6.1 Terra

GPT-6.1 Terra is an AI model released by OpenAI on Oct 26 2026. It has a 1.05M context window. Tracked results include 74.2% on OSWorld 2.1 (offline), 60% on Agent's Last Exam (pass@1) and 85.5% on MMMU-Pro.

## Facts

| Field | Value |
| --- | --- |
| Model | GPT-6.1 Terra |
| Developer | OpenAI |
| Release date | Monday, Oct 26 2026 |
| Licensing | Closed |
| Context window | 1.05M |

## API pricing

All rates in USD per 1,000,000 tokens, pay-as-you-go.

| Tier | Input | Cached input | Cache write | Output |
| --- | --- | --- | --- | --- |
| Up to 272K input tokens | $1.00 | $0.05 | $1.25 | $5.00 |
| Over 272K input tokens | $2.00 | $0.10 | $2.50 | $7.50 |

Verified October 19, 2026 against the first-party source: https://developers.openai.com/api/docs/models/gpt-6.1-terra

## Tracked benchmark scores

| Benchmark | Score | Source | What it measures |
| --- | --- | --- | --- |
| OpenAI Factuality (Internal) | 4.3% | Lab | OpenAI's internal test uses prompts designed to provoke factual mistakes. The score is the percentage of answers containing at least one factual error, at the model's best published reasoning effort. It is not the percentage of individual claims that are wrong or an error rate for everyday use. Lower is better. |
| Terminal-Bench 4.0 | 50.3% | Lab | Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better. |
| Terminal-Bench-Science 0.1 | 58.3% | Lab | The same command-line setup as Terminal-Bench, pointed at scientific work: the AI has to drive research tooling and computational workflows through to a result, rather than administer a machine. Version 0.1 is the first release of the task set, and scores run lower than on the general board. Higher is better. |
| OSWorld 2.1 (offline) | 74.2% | Lab | Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Version 2.1, in its offline setting. Scores are not comparable with version 2.0. Higher is better. |
| Agent's Last Exam (pass@1) | 60% | Lab | A hard set of desktop and operating-system tasks an AI agent has to finish by looking at the screen and working the machine itself. The score is the share it passes outright — partial credit does not count. Higher is better. |
| MMMU-Pro | 85.5% | Lab | A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better. |

## About GPT-6.1 Terra

GPT-6.1 Terra was introduced on October 26, 2026 as OpenAI's fast, efficient model for visual reasoning, computer use and writing, priced below GPT-6.1 Sol. Its documented uses included matching product photographs to catalogue entries, interpreting technical drawings and acting on application screens. OpenAI also positioned it for concise writing from supplied material and scoped code fixes, giving the GPT-6 family a Terra option for everyday work alongside Astra, Sol and Luna.

The API accepted text and image inputs and generated text, with tool calling through the Responses API. Reasoning effort ranged from none to maximum, with medium as the default, allowing developers to adjust effort to the task. At launch, OpenAI reported 85.5% on MMMU-Pro and 74.2% on the OSWorld 2.1 offline set at maximum effort. These were publisher-reported results under specific evaluation conditions; application performance also depended on the tools, instructions and reasoning setting used.

## Questions and answers

### When was GPT-6.1 Terra released?

GPT-6.1 Terra was released by OpenAI on Monday, Oct 26 2026.

### Who made GPT-6.1 Terra?

GPT-6.1 Terra was built by OpenAI. Creators of ChatGPT and the GPT series of models. Pioneered large-scale language model research.

### How much does GPT-6.1 Terra cost?

GPT-6.1 Terra costs $1.00 per million input tokens and $5.00 per million output tokens through the OpenAI API. Cached input is $0.05 per million tokens. Those are the rates for the “Up to 272K input tokens” tier; 1 other pricing tier is published for this model. Rates are pay-as-you-go API prices verified against OpenAI's published pricing on October 19, 2026.

### What benchmark scores did GPT-6.1 Terra get?

GPT-6.1 Terra reports 6 tracked benchmark scores — OpenAI Factuality (Internal): 4.3%; Terminal-Bench 4.0: 50.3%; Terminal-Bench-Science 0.1: 58.3%; OSWorld 2.1 (offline): 74.2%; Agent's Last Exam (pass@1): 60%; MMMU-Pro: 85.5%. Tracked scores may come from lab reports or independent benchmarks; source details accompany the benchmark data. It holds the best score among all models tracked here on OpenAI Factuality (Internal), OSWorld 2.1 (offline), Agent's Last Exam (pass@1) and MMMU-Pro.

### What is the context window of GPT-6.1 Terra?

GPT-6.1 Terra has a context window of 1.05M. Providers may specify separate input and output token limits.

### Is GPT-6.1 Terra open source?

No. GPT-6.1 Terra is a closed model. The weights are not published — it is available only through the provider's own API, apps, or partner platforms.

### What came before and after GPT-6.1 Terra?

OpenAI's previous tracked release was GPT-6.1 Sol on Sep 29 2026, 27 days earlier. It is the most recent OpenAI model tracked on AI Release Tracker.


---

Canonical page: https://aireleasetracker.com/model/openai/gpt-6.1-terra
Site index: https://aireleasetracker.com/llms.txt
Source: AI Release Tracker (https://aireleasetracker.com). Most benchmark scores come from lab launch material; gathered results identify the leaderboard that published them.