OpenAI has announced GPT‑6.1 Terra, a mid-range model the company highlights for its visual reasoning and computer-use capabilities. The release notes stress speed and efficiency, detailed visual interpretation, precise computer interactions, and clearer, more natural text generation. OpenAI describes GPT‑6.1 Terra as a notable upgrade over GPT‑5.6 Terra.
Positioning and use cases
Terra occupies a middle position in OpenAI's lineup: neither the cheapest nor the most expensive option, comparable in role to Anthropic's Sonnet models. That placement can make selection less straightforward, since some tasks—especially automated, repetitive computer tasks—may be better suited to lower-cost models.
In community commentary, Ted Sanders suggested that Terra may outperform Astra and Sol in technical communication, be similar on vision and computer use (but much faster), and be somewhat weaker on many other tasks. He framed Terra as an appealing midrange choice for API developers who want something cheaper than Sol but notably smarter than Luna, particularly for vision, computer use, and communication. For subscribers, he noted it is not a dramatic change; it may help those hitting usage limits or wanting clearer communication and faster computer actions.
Pricing and performance details
OpenAI lists GPT‑6.1 Terra at $1 per million input tokens and $5 per million output tokens. That pricing is roughly half of GPT‑5.6 Terra and GPT‑6.1 Sol, and about ten times the cost of GPT‑6 Luna.
The announcement included comparative materials and a performance example: a ‘‘pelican’’ measurement for the “max” reasoning level showed a cost of 11.3161 cents and took 4 minutes 11 seconds to run. A comparison grid provided pelican metrics across the current frontier models: GPT‑6 Astra, GPT‑6.1 Sol, GPT‑6.1 Terra, and GPT‑6 Luna.
Why it matters
GPT‑6.1 Terra targets daily work tasks that benefit from vision and computer interaction—drafting proposals, interpreting technical drawings, filling forms, and editing videos are all cited as representative applications. The mid-tier pricing aims to give developers and subscribers an option that balances capability and cost: better visual and automation features than entry-level models, without the top-tier price.
Summary
GPT‑6.1 Terra is positioned as a middle-ground model emphasizing visual reasoning and precise computer use. Its pricing and performance profile may make it a practical choice for API users and subscribers who need improved vision and automation features without paying for the most expensive models.
