GPT-6.1 Terra (source). It’s always interesting to see what gets highlighted in announcements for new models.
In the case of GPT-6.1 Terra, OpenAI focused on visual reasoning and computer use:
We’re launching GPT-6.1 Terra, a fast, efficient model that combines nuanced visual reasoning and precise computer use with clearer, more natural writing. A major step up from GPT-5.6 Terra, it excels at communication and computer-based workflows that fill the workday—from drafting proposals and interpreting technical drawings to form filling and video editing.
Terra occupies a similar spot in OpenAI’s lineup to Sonnet in Anthropic’s model family—it’s neither the cheapest nor the most expensive, which makes deciding when to reach for it a bit trickier.
Computer use is a classic example of a workload that benefits from lower-cost models, especially when setting them loose on repetitive automated tasks.
OpenAI’s Ted Sanders offered some practical advice in a comment on Hacker News:
For GPT-6.1 Terra, I’d say it’s probably better than Astra and Sol at technical communication; on par for vision and computer use (though much faster!); and slightly worse at most other things. For API developers, it’s a solid mid-tier option if you want something cheaper than Sol but noticeably smarter than Luna (especially for vision/computer use/communication). For subscribers, it’s not a game-changing update—it might stretch your rate limits a bit further, and it’s a great pick if you want clearer prose or faster computer actions.
Here are the pelicans generated across various reasoning levels. The “max” reasoning pelican cost 11.3161 cents and took 4 minutes and 11 seconds:

Here is a new comparison grid showing pelicans generated across reasoning levels for OpenAI’s current frontier lineup: GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna.
GPT-6.1 Terra is priced at $1 per million input tokens and $5 per million output tokens—roughly half the price of GPT-5.6 Terra and GPT-6.1 Sol, and about 10x the price of GPT-6 Luna.