GPT-6.1 Terra
OpenAI released GPT-6.1 Terra, a mid-tier model for computer use and writing, priced at $1 per million input and $5 per million output tokens, half the price of GPT-5.6 Terra.
Terra replaces GPT-5.6 Terra and adds stronger visual reasoning and computer use. On Agents' Last Exam at medium effort it scores 15 percentage points higher than GPT-5.6 Terra at about one-third the cost. Its error rate on error-inducing prompts at low effort fell from 53.6% to 16.7%.
- Date
- Monday 26 October 2026
- Lab
- OpenAI
- Kind
- model
- Access
- closed API
- Price
- $1 per million input tokens, $0.05 per million cached input tokens, $5 per million output tokens
Figures
| Measure | Value | Measured by |
|---|---|---|
| Terminal-Bench 4.0 at max effort | 50.3% GPT-5.6 Terra published score was 21.5% | company |
| Error rate on error-inducing prompts, low effort | 16.7% down from 53.6% for GPT-5.6 Terra | company |
| Error rate on error-inducing prompts, max effort | 4.3% down 18 percentage points | company |
| Agents' Last Exam gain over GPT-5.6 Terra, medium effort | +15 percentage points at about one-third the cost | company |
| Task speed | over 3x faster than GPT-6 Astra and GPT-6.1 Sol | company |
Available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and in the API as gpt-6.1-terra. It meets the Critical cybersecurity threshold under OpenAI's Preparedness Framework. Benchmark results are from OpenAI's own testing.
Sources
- openai.com/index/introducing-gpt-6-1-terra
- developers.openai.com/api/docs/changelog
- thenextweb.com/news/openai-gpt-6-1-terra-computer-use
This record was checked against its sources on 26 October 2026. How we check