Open Brain
VercelEngineering PostOfficial Source

GPT-6.1 Terra now available on AI Gateway

GPT-6.1 Terra is now available through Vercel AI Gateway for coding agents and API clients, with image input, tool calling, and a context window above 1M tokens.

Vercel · Oct 26, 2026
Open Source Open MarkdownOpen JSON
Source Summary

**GPT-6.1 Terra** is available on AI Gateway for the AI SDK, OpenAI-compatible Chat Completions, Responses API, and connected coding agents. It accepts text and images, returns text, and has a context window of **more than 1M tokens**.

Practical Implication

Consider it for high-volume document and desktop workflows involving visual reasoning or computer use. It supports configurable reasoning effort, structured outputs, tool calling, prompt caching, and a fast mode; the model name is `openai/gpt-6.1-terra`.

Agent-Ready Context
**GPT-6.1 Terra** is available on AI Gateway for the AI SDK, OpenAI-compatible Chat Completions, Responses API, and connected coding agents. It accepts text and images, returns text, and has a context window of **more than 1M tokens**.

Consider it for high-volume document and desktop workflows involving visual reasoning or computer use. It supports configurable reasoning effort, structured outputs, tool calling, prompt caching, and a fast mode; the model name is `openai/gpt-6.1-terra`.

The source says it improves on the previous Terra in visual reasoning, factual accuracy, and coding, but recommends larger GPT-6 models for the most complex coding tasks. No comparative measurements are provided.
Connected Context · Feed7 Judgment

Terra adds a gateway option for workloads needing image input and a context window above one million tokens, with configurable reasoning and tool use. The source offers provider claims but no comparative measurements, so it broadens the trial set without establishing a production model choice.

Context Map
modelcodingimage#model-selection#reasoning#computer-use
Uncertainty
The source says it improves on the previous Terra in visual reasoning, factual accuracy, and coding, but recommends larger GPT-6 models for the most complex coding tasks. No comparative measurements are provided.