OpenAI: o4 Mini

OpenAI o4 Mini is a compact reasoning model from OpenAI, part of the company's o-series. It was released by the provider on 16 April 2025, and its knowledge cutoff is 30 June 2024. The model sits in the text category and is positioned as a smaller sibling to OpenAI's larger reasoning models, aimed at fast and cost-efficient performance while keeping strong multimodal and agentic capabilities.

The model accepts text, images and files as input, and it returns text only. Its context window is 200,000 tokens, with a maximum output of 100,000 tokens. Because input can include images and files alongside text, you can send prompts that mix written instructions with visual or document content in a single request. The reasoning behaviour is exposed through the reasoning and include_reasoning parameters, so you can control whether reasoning output is included in the response.

Those capabilities point to a few practical patterns. Reasoning-heavy tasks such as multi-step analysis, planning and problem decomposition fit the model's design. Because it accepts images and files, it also suits workflows where a prompt needs to reference a screenshot, diagram or document rather than plain text alone. Tool use and structured outputs make it a reasonable choice for agent-style applications, for example pipelines that call functions, query external systems or need responses in a fixed schema. The large context window helps when a task depends on long inputs, such as extended documents or lengthy conversation histories.

You call the model through api14's OpenAI-compatible API using the model id "openai/o4-mini". The supported request parameters are include_reasoning, max_tokens, reasoning, response_format, seed, structured_outputs, tool_choice and tools. In practice this means you can cap output length with max_tokens, request deterministic sampling with seed, ask for structured or schema-constrained responses through response_format and structured_outputs, and define callable functions with tools and tool_choice. The reasoning and include_reasoning parameters let you manage how reasoning is handled in the response.

The main constraint to keep in mind is that output is text only, so tasks that require generated images, audio or other non-text output are outside what this model returns. It is also a compact model in the o-series, so if a task demands the strongest reasoning available, a larger model in the same family may be a better fit. For most everyday reasoning, multimodal input and tool-driven workloads, o4 Mini covers the ground with a 200,000 token context window and a 100,000 token output ceiling.