Dots Studio: Dots3-Note Preview (free)

Dots Studio: Dots3-Note Preview (free) is a text generation model from Dots Studio. It is an open-weight mixture-of-experts model with 16B active parameters out of 280B total, and it is the lightest model in the Dots 3 family. The provider release date is 2026-08-14. The model id on api14 is dots-studio/dots-3-note-preview:free.

The model accepts text and images as input and produces text as output, so it can handle prompts that combine written instructions with visual content. Its context window is 512000 tokens, and the maximum output is 460800 tokens. That combination of a large context and a large output budget makes it suitable for long documents and long-form responses rather than short single-turn exchanges.

Because it takes image input alongside text, practical uses include asking questions about screenshots, diagrams, charts or scanned pages, and working through documents where the visual layout matters. The large context window supports tasks such as summarising or extracting information from lengthy reports, codebases or transcripts, and the large maximum output allows for extended generated text in a single request. The model also supports reasoning and tool use, which fits workflows where the model needs to call functions or produce structured results rather than free-form prose.

You call it through api14's OpenAI-compatible API using the model id dots-studio/dots-3-note-preview:free. The supported request parameters are include_reasoning, max_tokens, reasoning, response_format, structured_outputs, temperature, tool_choice, tools and top_p. In practice this means you can control sampling with temperature and top_p, cap the response with max_tokens, request structured or JSON output through response_format and structured_outputs, and define callable functions with tools and tool_choice. The reasoning and include_reasoning parameters let you enable reasoning behaviour and control whether reasoning content is returned.

Since this is a preview release and the lightest model in its family, it may not be the right choice when you need the strongest reasoning or the largest model available. If a task demands capabilities beyond what a lightweight preview offers, another model in the Dots 3 family or a different provider may fit better. For image understanding, long-context document work and tool-driven text generation, it is a reasonable starting point to evaluate.