OpenAI: GPT-5.2 (batch)

GPT-5.2 (batch) is a text model from OpenAI, listed on api14 under the model id openai/gpt-5.2:batch. It belongs to the GPT-5 series, and OpenAI describes it as the latest frontier-grade model in that line, with stronger agentic and long context performance than GPT-5.1. The batch label indicates this is the batch variant of the model rather than the standard interactive endpoint.

The model accepts text, images and files as input, and it returns text only. Its context window is 400,000 tokens, which is large enough to hold long documents, extended conversations or sizeable codebases in a single request. The maximum output is 128,000 tokens, so responses can be long when the task calls for it. According to OpenAI, GPT-5.2 uses adaptive reasoning to allocate computation dynamically, which means it can respond quickly on simple prompts while spending more effort on harder ones.

Those capabilities point to a fairly specific set of use cases. The combination of image and file input with a 400,000 token context suits document analysis, where you pass in contracts, reports or PDFs alongside instructions, and summarisation or extraction over long source material. The agentic strength OpenAI claims makes it a reasonable fit for tool-driven workflows, since the model supports tools and tool_choice, and for tasks that need structured output rather than free text. Long context also helps with multi-file code review or question answering over a large internal corpus, where chunking would otherwise lose the thread.

You call it through api14's OpenAI-compatible API using the model id openai/gpt-5.2:batch, so existing OpenAI SDKs and clients work with only a base URL and key change. The supported request parameters are include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, seed, structured_outputs, tool_choice, tools and verbosity. In practice, reasoning_effort and verbosity let you trade latency and output length against depth, response_format and structured_outputs cover JSON-style responses, and seed helps when you want repeatable sampling. The reasoning and include_reasoning parameters expose the model's reasoning behaviour where your client can consume it.

One thing to weigh is that this is the batch variant, so it is aimed at workloads you can queue rather than latency-sensitive interactive chat. If you need immediate turn-by-turn responses, the non-batch version of the same model is the more natural choice. Output is text only, so image or audio generation is out of scope here.