Generative Pre-trained Transformer 6.1 Terra arrives.
OpenAI LLM Terra offers mid-tier pricing, visual reasoning, computer use, and natural writing for mixed tasks, while cheaper automation favors Luna.
Key Takeaways
- OpenAI introduced GPT-6.1 Terra, a fast and efficient model combining detailed visual reasoning, precise computer use, and clearer, more natural writing.
- Terra is positioned as a middle-tier model, about half the price of GPT-5.6 Terra and GPT-6.1 Sol, but roughly 10x the price of GPT-6 Luna.
- Its pricing is $1 per million input tokens and $5 per million output tokens, making cost trade-offs important for long prompts and automated workflows.
- Practical uses highlighted include drafting proposals, interpreting technical drawings, filling out forms, and editing videos.
- Hacker News pelican examples from Ted Sanders provide additional cost, time, and reasoning-level context across GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna.
Meta description: OpenAI introduces GPT-6.1 Terra, a fast model with visual reasoning and computer use, priced at $1 per million input and $5 per million output tokens.
Tags: ai, openai, generative-ai, llms, llm-pricing
Featured image: abstract friendly illustration of a pelican riding a bicycle beside visual reasoning and computer-use icons, no real people, no logos, no text.
OpenAI introduced GPT-6.1 Terra, a fast, efficient model that combines detailed visual reasoning and precise computer use with clearer, more natural writing. The release announcement emphasized those capabilities. OpenAI described Terra as a significant step up from GPT-5.6 Terra for communication and computer-based tasks that fill the workday, including drafting proposals, interpreting technical drawings, filling out forms, and editing videos.
Where Terra sits in the lineup
Terra is not the cheapest or the most expensive option. The source compares its position to Sonnet in the Anthropic model collection, meaning it occupies a middle tier. That placement makes it harder to decide when to use it, because users must judge whether a task needs Terra’s visual reasoning, writing quality, and computer-use precision instead of another model.
Computer use is a useful example. The source notes that it benefits from a less expensive model, since a user might put it to work on automated, repetitive tasks. Terra’s lower price compared with GPT-5.6 Terra and GPT-6.1 Sol may make that kind of workflow more attractive. At the same time, Terra costs around 10x the price of GPT-6 Luna, so the cheapest route remains Luna for many repetitive automation jobs.
The pricing numbers and the trade-off
GPT-6.1 Terra is priced at $1/million input tokens and $5/million output. That is almost half the price of GPT-5.6 Terra and GPT-6.1 Sol, and around 10x the price of GPT-6 Luna. These figures give readers a concrete way to compare options.
Because prices are quoted per million tokens, the difference can matter when prompts are long, when outputs are generated repeatedly, or when a workflow is automated. If a user is comparing Terra with GPT-5.6 Terra and GPT-6.1 Sol, the new model may be worth evaluating for its lower price. If a user is comparing Terra with GPT-6 Luna, Luna may remain the lower-cost option when price is the main constraint.
Ted Sanders’ Hacker News pelican examples
OpenAI’s Ted Sanders provided useful advice in a comment on Hacker News. He shared pelicans for the different reasoning levels, and noted that the pelican for "max" cost 11.3161 cents and took 4 minutes and 11 seconds. He also posted a new comparison grid showing pelicans for every reasoning level of OpenAI’s current frontier lineup: GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna.
The examples help readers connect model choice with cost and time. The 11.3161 cents and 4 minutes and 11 seconds figure shows that a max reasoning-level pelican task can carry measurable cost and duration. For workday tasks that are automated or repetitive, that matters because many requests may run over time. The comparison grid also places Terra among the current frontier lineup, giving side-by-side context against GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna.
What the release means for practical tasks
The capabilities emphasized in the release map directly to practical work. Drafting proposals benefits from clearer, more natural writing. Interpreting technical drawings and filling out forms depend on detailed visual reasoning and precise computer use. Editing videos points to support for creative and computer-based tasks.
For readers, GPT-6.1 Terra is most interesting when a workflow needs the combination described in the release but does not require the most expensive option. It may be a candidate for mixed tasks that involve language, visual understanding, and interaction with software. The pricing, $1/million input tokens and $5/million output, makes that middle tier easier to evaluate, while the Hacker News pelican grid gives additional context for reasoning levels across GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna.

Comments (0)