GPT 6.1 Terra

Supported Functionality

Item	Specification
Input	Text, images
Output	Text
Context	1,050,000 tokens
Max Output	128,000 tokens
Vision	✓ Supported
Function Calling	✓ Supported; tool calling requires the Responses API

Description

GPT-6.1 Terra is an efficient reasoning model released by OpenAI on October 26, 2026. Its official API identifier is gpt-6.1-terra. It accepts text and images, produces text, supports a 1,050,000-token context window and up to 128,000 output tokens, and has a June 1, 2026 knowledge cutoff. Direct audio and video input are not supported.

OpenAI positions the model for visual reasoning, precise computer use, and clear, natural writing. Representative tasks include matching product photos to catalog entries, interpreting technical drawings, acting on application screens, and drafting or revising text from supplied material. Its reasoning.effort accepts none, low, medium, high, xhigh, and max, with medium as the default. Tool calling and built-in tools require the Responses API.

Key Capabilities

• Image and Drawing Understanding: Examines product photos, technical drawings, charts, and screenshots to extract relevant visual details.
• Precise Computer Use: Acts from application screen state in workflows with a defined goal and verifiable outcome.
• Natural Writing: Drafts, revises, and condenses supplied material while preserving facts, tone, and readability.
• Scoped Code Changes: Understands focused code requirements, makes targeted fixes, and supports test-based validation.
• Long-Context Processing: Connects large document sets, catalogs, interface records, and tool results in one request.
• Tiered Reasoning: Offers six effort levels from none to max for evaluation across task difficulty and response needs.
• Tool Collaboration: Uses the Responses API with function calls, search, Code Interpreter, Computer Use, MCP, and other supported tools.

Technical Strengths

Feature	Benefit
1.05M-Token Context	Holds large catalogs, lengthy source material, and interaction history with less segmentation.
128K Output Limit	Supports longer reports, structured results, and substantial code or document output.
Six Reasoning Levels	Adjusts effort from none to max, defaulting to medium, for different visual, writing, and action tasks.
Text and Image Input	Combines visual evidence with written requirements for matching, judgment, and screen understanding.
Responses API Tooling	Supports functions, web search, file search, Code Interpreter, Computer Use, MCP, and other tools.
Streaming and Structured Outputs	Enables live responses and structured data for interactive interfaces and downstream systems.

Frequently Asked Questions

What input and output modalities does GPT-6.1 Terra support?

It accepts text and images and generates text; the official model page does not list audio or video as supported input modalities. It can read photos, drawings, and application screenshots, while image generation is a separate tool capability invoked through the Responses API rather than native image output.

How do I verify a GPT-6.1 Terra computer-use task?

Define the goal, permitted actions, and completion criteria before execution, then inspect the saved record or actual interface state. A completion statement alone is insufficient; keep screenshots, logs, or business-system evidence for important actions.

Should I choose GPT-6.1 Terra or GPT-6.1 Sol?

Terra fits visual reasoning, defined screen tasks, concise writing, and scoped code fixes. For complex engineering, professional projects, or more demanding long-running work, compare it with Sol on identical inputs and acceptance criteria, evaluating quality, tool success, and observed latency.

What long-context work suits GPT-6.1 Terra?

Its 1.05M-token window can accommodate large product catalogs, document collections, extended interaction history, and related image descriptions. Identify source boundaries, task priorities, and output requirements, then verify important facts against the original material.

Where do I find GPT-6.1 Terra's LinkModel calling details?

Check LinkModel's live model directory and its model detail page for the public model ID, available endpoints, and request fields. OpenAI's gpt-6.1-terra is the first-party API identifier; use LinkModel's current catalog to confirm its own route and supported features.

How do I validate GPT-6.1 Terra integration on LinkModel?

Use the model ID and endpoint displayed on LinkModel's current model page for a basic text request, then test image input, reasoning effort, long context, structured output, and function calls independently. Record response fields, errors, truncation, and tool results, and configure only capabilities that the LinkModel page confirms.