GPT-6.1 Terra
openai/gpt-6.1-terra
Back to model catalog

OpenAI · Model guide

GPT-6.1 Terra

Compare
openai/gpt-6.1-terraReleased Oct 26, 2026

A fast, efficient GPT-6.1 model for visual reasoning, precise computer use, and clear, natural writing. It accepts text and images, supports a 1.05M-token context window, and offers reasoning effort from none to max.

Modalities
TextImageText
Context
1M
Max output
128K
Price / 1M tokens
$ 1.00in$ 5.00out
StreamingTool callingJSON outputImage inputThinking mode

What this model is good at

What the vendor positions it for, and which of our models to reach for instead.

VendorA fast, efficient model for visual reasoning, precise computer use, and clear, natural writing.source

Best for
  • Visual reasoning · Match product photos to catalog entries and interpret technical drawings.(vendor claim)source

  • Computer use · Act precisely on application screens.(vendor claim)source

  • Writing · Draft and revise clear, natural text.(vendor claim)source

Neighbouring models
Compare them side by side →

Pricing and billing

Tiered pricing: the rate changes once the input passes the threshold.

Price / 1M tokens by input sizeInput < 272.0KInput ≥ 272.0K
Input$ 1.00$ 2.00
Output$ 5.00$ 7.50
Cache read$ 0.050$ 0.10
Cache write$ 1.25$ 2.50
Estimate a request

At this input size you are on the Input < 272.0K tier.

$ 0.0200
Estimated cost per request
10.0K × $ 1.00
+ 2.0K × $ 5.00
≈ $ 20.00 per 1,000 requests

Capabilities and limits

What CrossModel guarantees across every route this model can take right now.

Context window
1.1M tokens
Max output
128.0K tokens
Input / output modalities
Text + Image → Text
Streaming
Supported
Tool calling
Supported
Structured output (JSON)
Supported
Image input
Supported
Thinking mode
Supported· always on
Reasoning effort
none · low · medium · high · xhigh · max
Available endpoints
/v1/chat/completions · /v1/responses · /v1/messages

Published benchmarks

Scores the vendor reported, with the evaluation setup each one came from.

Agentic coding
  • Terminal-Bench 4.0
    pass_rate·OpenAI · 2026-10-26effort maxsetup not fully disclosed

    官方发布页交互图表(2026-11-10 核对)各推理档位最高成绩,明确记录对应 effort。发布页正文:最高推理强度下解决 50.3% 的任务。harness 未披露。完整复现设置未披露;研究/API 配置可能不同于生产。

    50.3%
Agentic work
  • Agents' Last Exam
    score·OpenAI · 2026-10-26effort maxsetup not fully disclosed

    官方发布页交互图表(2026-11-10 核对)各推理档位最高成绩,明确记录对应 effort。图表标题为 Agents’ Last Exam v1;同图 GPT-6 Astra Max 59.3 与目录既有成绩一致,故沿用该身份。专业任务中的电脑操作(如建筑图纸转 3D 模型)。完整复现设置未披露;研究/API 配置可能不同于生产。

    60%
Computer use
  • OSWorld 2.1 (offline subset, partial score)
    score·OpenAI · 2026-10-26effort maxsetup not fully disclosed

    官方发布页交互图表(2026-11-10 核对)各推理档位最高成绩,明确记录对应 effort。图表标题为「OSWorld 2.1, 离线测试集」;同图 Claude Sonnet 5.5 Max 83.9 与 Anthropic 在离线子集配置下的重测值一致,故归入此身份。不是 OSWorld 2.1 全集或 OSWorld-Verified。完整复现设置未披露;研究/API 配置可能不同于生产。

    74.2%
Cybersecurity
  • ExploitBench
    score·OpenAI · 2026-10-26setup not fully disclosed

    System card 9.0.2.1 节的 Preparedness 网络安全评测,ACE-normalized mean capability score;按最大测试解长度报告,未披露推理档位。OpenAI 自述历史漏洞可能带来训练数据污染、使成绩虚高。研究/API 配置可能不同于生产。

    97.1%
Knowledge work
  • HealthBench Professional
    score·OpenAI · 2026-10-26setup not fully disclosed

    System card 5.1 节 HealthBench 表的 length-adjusted 分数;未做长度调整为 64.2,平均回答长度 2531 字符。同表 GPT-6.1 Sol 64.2、GPT-5.6 Terra 57.7。该节未披露推理档位;研究/API 配置可能不同于生产。

    62.7%
Multimodal reasoning
  • MMMU Pro (vision)
    score·OpenAI · 2026-10-26effort maxsetup not fully disclosed

    官方发布页交互图表(2026-11-10 核对)各推理档位最高成绩,明确记录对应 effort。图表标题为「MMMU-Pro (视觉)」,即题目嵌入图片的 vision 设定;页面未说明是否带工具,因此不与目录中 with-tools / no-tools 合并(同图 GPT-5.6 Terra Max 81.7,而 GPT-5.6 发布页 with-tools 为 82.0)。完整复现设置未披露;研究/API 配置可能不同于生产。

    85.5%
Scientific research
  • Terminal-Bench-Science 0.1
    accuracy·OpenAI · 2026-10-26effort maxsetup not fully disclosed

    官方发布页交互图表(2026-11-10 核对)各推理档位最高成绩,明确记录对应 effort。同图 GPT-6.1 Sol Max 57.0 与目录既有成绩一致。不同于 Terminal-Bench 2.x/4.0。完整复现设置未披露;研究/API 配置可能不同于生产。目录沿用此基准既有 accuracy 指标名;基准官网称 resolution rate(任务解决比例)。

    58.3%

Vendor-reported numbers, not CrossModel measurements. Scores are only comparable when the benchmark version, metric and evaluation setup match, so nothing here is averaged or ranked.

Use it in your tools

Point the base URL at CrossModel and paste this model ID — every tool below has a setup guide.

See all integration guides →

Frequently asked questions

What is GPT-6.1 Terra?
GPT-6.1 Terra is available on CrossModel as openai/gpt-6.1-terra. It has a 1M-token context window and can return up to 128K tokens per request.
How much does GPT-6.1 Terra cost?
Currently $ 1.00 per 1M input tokens and $ 5.00 per 1M output tokens. Cached input is billed at $ 0.050 per 1M.
Does GPT-6.1 Terra support tool calling and structured output?
Tool calling: Supported · Structured output (JSON): Supported · Image input: Supported · Streaming: Supported · Thinking mode: Supported
Which endpoint do I call?
For text models, OpenAI Chat Completions (/v1/chat/completions), OpenAI Responses (/v1/responses) and Anthropic Messages (/v1/messages) all work — pick whichever your SDK already speaks.
Can I try it without writing code?
Yes. Sign in and open the Playground on this page to chat with the model directly — no API key needed. Usage is billed from your wallet at the normal rate.

Chat first, integrate later

Chat first, then wire it in once you like the answers.