GPT-5.1
A five-model GPT family with instant, reasoning and coding variants
GPT-5.1 is a set of five large language models from OpenAI, part of its GPT line. Three arrived on 12 November 2025 and the remaining pair a week later, replacing GPT-5 and GPT-5 Pro while adding customisable personalities and dedicated coding and reasoning modes.
- Five models in the family, released across two dates in November 2025
- Eight selectable personalities, with a warmer default tone than GPT-5
- Two modes: Instant for everyday use, Thinking for harder reasoning
- Codex-Mini and Codex-Max cover coding and agentic tasks
- Proprietary models included with ChatGPT and Microsoft Copilot
What it is
GPT-5.1 is a group of five large language models released by OpenAI as part of its GPT series. The first three became available on 12 November 2025, with two further models following on 19 November. The family covers general chat, harder reasoning work and coding or agentic tasks, and it is multimodal and generative pre-trained transformer software. It takes over from GPT-5 and GPT-5 Pro and is included with ChatGPT and Microsoft Copilot. The weights are proprietary, so users reach the models through OpenAI's products rather than by downloading them.
How it came about
OpenAI published GPT-5.1 on 12 November 2025. That day brought GPT-5.1 Instant, GPT-5.1 Thinking and GPT-5.1-Codex-Mini. A week later, on 19 November, the company added GPT-5.1-Codex-Max, an agentic coding model, and GPT-5.1 Pro, which took over from GPT-5 Pro. Paid Pro, Plus, Go and Business subscribers were offered earlier access, while the models were opened to every user. GPT-5.2 later succeeded the family.
Inside the model
The family runs in two modes. GPT-5.1 Instant handles ordinary requests, while GPT-5.1 Thinking is a reasoning model aimed at harder problems. GPT-5.1-Codex-Mini was positioned as a cheaper option for coding and agentic work, and GPT-5.1-Codex-Max extends that line for agentic coding. The models are described as multimodal, large language models and generative pre-trained transformers. OpenAI also gave GPT-5.1 a set of eight selectable personalities and made its default tone warmer than that of GPT-5.
What it can do
Amanda Caswell, writing for Tom's Guide, rated the model highly for mathematics and coding yet judged Gemini 3 stronger overall. Eric Hal Schwartz of TechRadar compared it with GPT-5 and reported steadier obedience to instructions, while describing the step forward as incremental. Customisable personalities, with eight settings, let users adjust how the assistant responds.
How to use it
Everyone can use GPT-5.1, and paying Pro, Plus, Go and Business customers received priority access. The models ship inside ChatGPT and Microsoft Copilot. The licence is proprietary, so no weights are published for download.
Reception and debate
Reviews were mixed but broadly positive. Tom's Guide highlighted strong maths and coding performance, though its reviewer preferred Gemini 3 overall. TechRadar found the model more reliable at following instructions than GPT-5, but characterised the release as an incremental improvement rather than a large leap.
Why it matters
As the successor to GPT-5, GPT-5.1 continued OpenAI's GPT line and was later replaced by GPT-5.2. It introduced selectable personalities to the series and split the offering into instant, reasoning and coding variants, including agentic coding through the Codex models.
Questions people ask about GPT-5.1
What is GPT-5.1?
It is a family of five large language models from OpenAI, released in November 2025 as part of the GPT series. It replaces GPT-5 and GPT-5 Pro and covers general chat, reasoning and coding work.
How do the different models in the family differ?
GPT-5.1 Instant handles everyday requests and GPT-5.1 Thinking is a reasoning model for harder tasks. GPT-5.1-Codex-Mini offers a cheaper coding and agentic option, GPT-5.1-Codex-Max targets agentic coding, and GPT-5.1 Pro replaced GPT-5 Pro.
How can I use it?
The models are available to all users through ChatGPT and Microsoft Copilot, with paid Pro, Plus, Go and Business tiers given advanced access. The licence is proprietary, so the weights are not published.
What changed compared with GPT-5?
OpenAI gave GPT-5.1 a warmer default personality and added eight personality options. TechRadar found it more consistent at following prompts than GPT-5, though its reviewer called the release an incremental improvement.
How did reviewers rate it?
Tom's Guide praised its maths and coding, but its reviewer judged Gemini 3 better overall. TechRadar noted steadier prompt following than GPT-5 while viewing the gains as incremental.
Good reading and sources
- Official websiteOfficial site
- GPT-5.1: A smarter, more conversational ChatGPTOpenAI, Article
- OpenAI says the brand-new GPT-5.1 is ‘warmer’ and has more ‘personality’ optionsThe Verge, Article
The facts on this page come from the Wikipedia article "GPT-5.1", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryLLM rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.