New to this? Read it in simple words
- OpenAI’s new GPT-6.1 Terra costs $1 per million input tokens, down from $2.
- An outside tester scores it 48, ranking it 25th of 221 models.
- It scores 6 points more than the old Terra at less than a third of the cost per task.
- OpenAI’s claims about speed and its biggest model are not checked yet.
- Token
- A small piece of text, about three quarters of a word. AI companies charge by the million tokens.
- API
- A way for programs to use a service directly. Developers pay for each use instead of a monthly plan.
- Artificial Analysis
- An independent company that runs the same tests on many AI models and compares their scores and costs.
OpenAI calls GPT-6.1 Terra a fast, efficient model for everyday computer work. An outside tester has already measured it.
What is new
OpenAI added GPT-6.1 Terra to its API on October 26, its changelog shows. It describes the model as fast and efficient, with visual reasoning, precise computer use and clear, natural writing.
The standard price is $1 per million input tokens and $5 per million output tokens. Cached input, text the model has already seen, costs $0.05 per million. GPT-5.6 Terra cost $2 and $12, Artificial Analysis notes.
OpenAI says the model is available to all Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex. OpenAI recommends its larger GPT-6.1 Sol and GPT-6 Astra for the hardest coding tasks.
From Artificial Analysis’s tests of October 26, 2026.
What OpenAI claims
OpenAI says GPT-6.1 Terra performs on par with GPT-6 Astra on Agents’ Last Exam, a test of professional computer tasks, at one-tenth of its price per token. It also says Terra completes tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol.
On OpenAI’s own test of prompts that invite mistakes, the error rate at low effort fell from 53.6% to 16.7%, the company says. Nobody outside OpenAI can run that test.
OpenAI says the model meets the “Critical” level for cybersecurity in its own safety framework. It keeps the protections built for GPT-6 Astra, the company says.
What the outside test found
Artificial Analysis, which tests AI models independently, scores GPT-6.1 Terra 48 on its Intelligence Index at maximum effort. That put it 25th of 221 models on October 27.
At maximum effort it scores 6 points more than GPT-5.6 Terra, at less than a third of the cost per task: $0.40 against $1.40, the tester says. It is still 4 points behind GPT-6.1 Sol.
On the tester’s own knowledge test, the model guessed in 38% of the cases where it did not know the answer. GPT-5.6 Terra guessed in 88%.
Sources
Every fact in this story comes from the sources below. Open them to check our work.
- 1
- 2
- 3
- 4Research · October 26, 2026OpenAI launches GPT-6.1 Terra with 3x faster AI agents, 50% lower pricing BleepingComputer
- 5Research · October 26, 2026OpenAI Launches GPT-6.1 Terra, Its Half-Good AI Model at Half the Price Decrypt
- 6Research · October 26, 2026GPT-6.1 Terra halves OpenAI’s API prices and still gains ground on its larger models The Decoder
The prices and the score of 48 are checked; the claims about GPT-6 Astra, speed and errors are OpenAI’s own. On one coding test, OpenAI’s gain is larger than the one Artificial Analysis measured, so we leave that number out.
OpenAI’s GPT-6.1 Terra costs less and scores higher. Silicon AI News, October 27, 2026. https://siliconainews.com/stories/openai-gpt-6-1-terra
OpenAI’s new GPT-6.1 Terra cuts its API prices, and an independent tester confirms it scores higher at under a third of the cost. How we rate the mood
The mood shows how the news frames the event for the tech and AI world: as an opportunity (6 and above) or as a loss (below 5). It is not our verdict on who is right.