OpenAI: GPT-5.2

GPT-5.2 is a text model from OpenAI and the latest frontier-grade entry in the GPT-5 series. It was released by the provider on 10 December 2025 and follows GPT-5.1, with stronger agentic and long context performance than that earlier version. It is a general purpose text model rather than a specialist tool, so it is aimed at developers who want one model to handle a wide range of language tasks through a single API.

The model accepts text, images and files as input, and it returns text. Its context window is 400,000 tokens, and the maximum output is 128,000 tokens, which means you can supply very large documents or long conversation histories and still leave room for a substantial response. GPT-5.2 uses adaptive reasoning, allocating computation dynamically so that it can respond quickly to simple requests while spending more effort on harder ones. It also supports reasoning controls, so you can influence how much reasoning work the model performs.

Those capabilities point to a few practical uses. Long context handling suits tasks such as summarising or querying lengthy documents, contracts, codebases or research material in a single request. Image and file input make it useful for extracting information from screenshots, scanned pages or uploaded documents and turning that content into text. The agentic strengths and tool support fit workflows where the model needs to call functions, follow multi step instructions or operate as part of a pipeline. Structured outputs and response format controls help when you need machine readable results rather than free prose.

You can call GPT-5.2 through api14's OpenAI-compatible API using the model id "openai/gpt-5.2". The request parameters it supports are include_reasoning, max_completion_tokens, max_tokens, reasoning, reasoning_effort, response_format, seed, structured_outputs, tool_choice, tools and verbosity. Because the interface follows the OpenAI format, existing OpenAI client libraries and request shapes should work with only the model id and base URL changed.

Keep the output ceiling in mind: responses are capped at 128,000 tokens even though the context window is larger, so very long generations may need to be split. If your task is purely text and does not need image or file input, reasoning controls or tool calling, a simpler model may be a better fit.