GPT-6.1 Terra
ONE-TENTH ASTRAReleased Oct 26, 2026; the mid-tier of OpenAI's GPT-6.1 series for visual reasoning, precise computer use (computer_use tool over app screens), and clear natural writing: $1/$5 per 1M tokens with cached input at $0.05 (5% of list) - half of GPT-6.1 Sol and GPT-5.6 Terra, one-tenth of GPT-6 Astra; 1,050,000-token context, 128,000 max output, reasoning effort none-max, batch and Flex at 50% off, and >272K prompts billed at 2x input / 1.5x output; live on the API, Codex, ChatGPT Work, GitHub Copilot, Amazon Bedrock, and OpenRouter; also served as gpt-6.1-terra-pro (reasoning.mode: pro).
Same underlying model served with reasoning.mode set to pro for higher-quality responses on complex tasks; pro mode spends more reasoning tokens. Also exposed as gpt-6.1-terra:batch for batch requests.
$1 in / $5 out per 1M tokens (same base rates; more reasoning tokens)
OFFICIAL DOCScurl -s https://modelregistry.tirup.in/api/cli?model=gpt-6-1-terraSources
- OpenAI API changelog (Oct 26, 2026: Released GPT-6.1 Terra) · OpenAI
- GPT-6.1 Terra model page · OpenAI
- OpenAI API pricing (GPT-6.1 Terra row) · OpenAI
- GPT-6.1 Terra listing and pricing · OpenRouter
- Codex model catalog · OpenAI
- Supported AI models in GitHub Copilot (GPT-6.1 Terra, GA) · GitHub
- Amazon Bedrock OpenAI model card: GPT-6.1 Terra · AWS