OpenAI releases GPT-6.1 Terra for agent workflows at half the API price OpenAI

OpenAI launched GPT-6.1 Terra, a mid-tier reasoning model optimized for agent workflows, computer use, and everyday coding, priced at 50% below GPT-5.6 Terra.

The model is available in ChatGPT Work, Codex, and via the API under the model identifier gpt-6.1-terra. In ChatGPT, rollout is underway for Plus, Pro, Enterprise, and Edu subscribers.

On internal benchmarks, OpenAI reported that GPT-6.1 Terra achieved a 50.3% solve rate on Terminal-Bench 4.0 at maximum reasoning effort, compared to 21.5% for GPT-5.6 Terra. On Agents' Last Exam, performance increased by 15 percentage points at medium reasoning effort while reducing task cost threefold. In evaluations testing resilience against adversarial prompts, the factual error rate dropped to 4.3%.

What it is

• Agent optimization: Designed specifically for multi-step agent execution, environment tool calls, and software engineering tasks.
• Terminal-Bench performance: Resolves 50.3% of terminal tasks at high reasoning effort, more than doubling predecessor accuracy.
• Lower hallucination rate: Hallucination and factual error frequency fell to 4.3% under complex adversarial evaluation.
• Broad availability: Accessible immediately via the OpenAI API as gpt-6.1-terra and rolling out across paid ChatGPT tiers.

Price

Sources

• Introducing GPT-6.1 Terra (OpenAI)