Writer: Palmyra X5

Writer: Palmyra X5 is a text model from Writer, listed on api14 under the model id writer/palmyra-x5. Writer describes it as its most advanced model, purpose-built for building and scaling AI agents across the enterprise. It sits in the text category, so it is aimed at language work rather than image, audio or other modalities.

The model takes text in and returns text out. Its context window is 1,040,000 tokens, which is just over a million tokens, and the maximum output is 8,192 tokens. Writer's own description highlights industry-leading speed and efficiency on context windows up to 1 million tokens, so the large window is a deliberate part of the design rather than an incidental figure. There is no vision or audio input here; everything you send must be text.

That combination of a very large context window and a modest output ceiling shapes the practical use cases. Long documents, large codebases, extended transcripts and multi-document sets can be supplied in a single request without chunking, which suits retrieval-heavy agent workflows, summarisation over long sources, question answering grounded in big reference material, and extraction tasks where the answer is short but the evidence is spread across many pages. Because the output is capped at 8,192 tokens, it fits best when you need a focused response rather than a long generated document.

You call it through api14's OpenAI-compatible API using the model id writer/palmyra-x5, so existing OpenAI client libraries and request shapes work without special handling. The supported request parameters are max_tokens, stop, temperature, top_k and top_p. Use max_tokens to bound the response within the 8,192 output ceiling, stop to define where generation should halt, and temperature, top_k and top_p to control sampling. No other parameters are listed as supported, so keep requests to this set.

The main constraint to plan around is the output limit of 8,192 tokens. If your task is to produce a long report, a full file or an extended piece of prose in one pass, this model will truncate before you are done, and a model with a larger output allowance may fit better. Palmyra X5 is strongest when the input is large and the required answer is comparatively small.