OpenAI has released the GPT-6.1 Terra model, focused on visual analysis, computer interaction, and everyday professional tasks. The new product is already available to users on the Plus, Pro, Enterprise, and Edu plans in the ChatGPT Work and Codex environments.
The key factor for the market was the pricing policy: the standard API cost is set at $1 per 1 million input tokens and $5 per 1 million output tokens. This is 50% lower than GPT-5.6 Terra and GPT-6.1 Sol, making the model one of the most cost-effective in the lineup for enterprise scenarios.
Accuracy and Performance
Internal tests demonstrate significant progress in combating hallucinations. At low reasoning effort, the share of factual errors dropped from 53.6% to 16.7% — more than threefold. On the Terminal-Bench 4.0 benchmark at maximum effort, the model solved 50.3% of tasks versus 21.5% for GPT-5.6 Terra, confirming a qualitative leap in agentic and terminal scenarios.
The positioning of GPT-6.1 Terra as a tool for visual analysis and computer work points to a shift in OpenAI's focus toward practical workflows, where not only reasoning but also execution accuracy at low inference cost matters.
Cryptalist analysis: Halving the API price while simultaneously reducing errors creates pressure on competitors in the enterprise AI solutions segment and accelerates the commoditization of inference. For crypto and Web3 projects building agentic services on top of LLMs, this means rethinking product economics toward cheaper and more reliable pipelines.