OpenAI launched GPT-6.1 Terra, enhancing computer operation, visual
understanding and coding capabilities. The model completes computer-operation
tasks more than three times faster than the prior generation and OpenAI reduced
API prices by 50%. API rates are $1 per mln input tokens, $0.05 per mln cached
input tokens and $5 per mln output tokens. Internal tests show low-inference
error rate fell from 53.6% to 16.7%; top-tier inference error rate is 4.3%, down
18 percentage points versus the previous generation. In coding tests, top-tier
scores rose from the prior published 21.5% to 50.3%. The model is available to
ChatGPT Work and CODEX Plus, Pro, Enterprise and Edu users and via API.