You can now test upcoming GPT-5.1 on OpenRouter

OpenAI is preparing to introduce a family of GPT-5.1 models, with references to GPT-5.1, GPT-5.1 Reasoning, and GPT-5.1 Pro recently discovered in the platform’s codebase. This approach suggests OpenAI is maintaining a similar product strategy as seen with GPT-5, continuing its trend toward simplified and transparent versioning. The code updates, specifically within Enterprise Permissions and Roles, indicate that enterprise customers will have the option to opt out of new model releases if they are considered experimental, a feature already familiar to larger organizations managing production workloads.

The latest code also references a general availability date of November 24th for the GPT-5.1 family. While this timeline is likely targeting enterprise accounts, there is a possibility that individual Plus and Pro users might gain access to the new models even earlier, possibly within the upcoming week. This timing appears to be aligned with broader industry trends, as OpenAI faces increased competition from releases like Kimi K2 Thinking and the anticipated Gemini 3 from Google.

There are also signs that one of the models, potentially GPT-5.1 Thinking, is already being tested under the name Polaris Alpha on platforms such as OpenRouter.

Community feedback points to high performance in creative writing and benchmark tests, increasing speculation about its true origin. If confirmed, this would suggest that public access to these capabilities is imminent.

For OpenAI, the staged rollout of the GPT-5.1 lineup on ChatGPT reinforces its strategy of segmenting models for specialized use cases and customer segments. These developments follow OpenAI’s established pattern of incremental but public-facing improvements to their core models, directly responding to evolving demands from both enterprise customers and the broader AI research community.