OpenAI rolls out GPT-5.1: Instant gets warmer, Thinking adjusts speed - here's who gets it first

What's new in GPT-5.1

• Models: GPT-5.1 Instant and GPT-5.1 Thinking.
• Instant: ChatGPT's most-used model now defaults to a warmer, more conversational tone; improves instruction following for direct answers; uses adaptive reasoning on challenging prompts while maintaining speed.
• Thinking: Adjusts its thinking time by task complexity. On the fastest tasks it runs about twice as fast as GPT-5 Thinking; on the slowest tasks it runs at about half the speed of GPT-5 Thinking.
• Responses use less jargon and fewer undefined terms, per OpenAI.

Personalization and tone controls

• Updated presets: Default, Friendly, and Efficient remain with refinements.
• New presets: Professional, Candid, and Quirky.
• Additional options available under the personalization menu: Cynical and Nerdy.
• Experimental sliders adjust conciseness, warmth, scannability, and emoji frequency.
• Personalization changes now apply across all chats immediately, including ongoing conversations.
• Models follow custom instructions more closely, according to OpenAI.

Rollout and availability

• Rollout begins for paid subscribers, then expands to free and logged-out users.
• Enterprise and Education admins have a seven-day early access toggle for GPT-5.1. The toggle is off by default.
• After seven days, GPT-5.1 becomes the default ChatGPT model.
• Prior models GPT-5 Instant and GPT-5 Thinking will remain in the legacy dropdown for three months for paid users.

Background

OpenAI positions Instant as optimized for speed and Thinking for advanced reasoning. The company compares GPT-5.1 Thinking with GPT-5 Thinking on representative ChatGPT tasks and says Instant's updates focus on delivering direct answers while keeping responses clear and concise. OpenAI also notes a phased rollout and a temporary retention period for prior models.