OpenAI releases GPT-6.1 Terra for cost-efficient computer use
OpenAI launched GPT-6.1 Terra, a mid-sized model priced at 50% of GPT-5.6 Terra, targeting high-volume automated tasks. Benchmarks show it offers strong cost-efficiency for computer use agents, though it trails higher-end models in agentic coding.
OpenAI released GPT-6.1 Terra, positioning it as a balanced option for speed, capability, and cost. The model is available at 50% the cost of GPT-5.6 Terra and half the cost of GPT-6.1 Sol. On the Agent's Last Exam benchmark, Terra scored slightly higher than other OpenAI models, including Astra, while completing tasks at half the cost of Sol and one-tenth the cost of Astra. It also approached the score of Opus 5.5 with three times the token efficiency. Artificial Analysis assigned Terra a score of 47, placing it four points below GPT-6.1 Sol and two points above Grok 4.7. On the AA Coding Agent Index, Terra scored 48, ranking behind GLM-53 but ahead of DeepSeq v4.1 Flash.
Developers and analysts suggest Terra is suitable for computer use tasks and high-volume automated workloads, such as invoice processing, where cost efficiency is paramount. However, the model scored slightly lower than Sol in agentic coding tasks, indicating it may not be the optimal choice for complex code production. OpenAI stated that Terra advances the cost-intelligence frontier, providing a new balance for users who require reliable performance without the premium cost of frontier models.