{"version":"2.1.288","anchor":"model-that-rejects-whole-pdfs-now-gets-a-read-pages-hint","canonical_anchor":"model-that-rejects-whole-pdfs-now-gets-a-read-pages-hint","heading":"Clearer PDF guidance for models that cannot take a whole PDF","tier":"notice","area":"Attachments","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.288\/e\/model-that-rejects-whole-pdfs-now-gets-a-read-pages-hint","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.288","markdown":"### Clearer PDF guidance for models that cannot take a whole PDF\n\nWhen a model cannot accept a whole PDF, Claude is now told to read specific pages with the Read tool's pages option instead of retrying\n\n**Unclear.** Which models trigger this message is not stated.\n\n**What**\n\nSome models cannot be sent a PDF file as a whole, only its pages as images. Two messages Claude sees now account for this:\n\n- When a document is removed because the model does not accept whole PDFs, the replacement text now says so (\"[Document removed: this model does not accept whole PDF documents.\") and tells Claude to read specific pages using the `pages` option of the Read tool, and not to re-read the file without `pages`. Before, a generic \"Document\/Image removed\" message was used.\n\n- The reminder for PDFs that are too large has a new version for these models. It says the model cannot be sent a PDF file, only its pages as images, suggests `pages: \"1-N\"` to read the whole thing when the page count is known, and uses \"page\" or \"pages\" correctly. Before, it always said \"This PDF is too large to read all at once.\"\n\nThe Read tool's own description is unchanged in this release.\n\n**Why**\n\nOn models without whole-PDF support, Claude gets accurate instructions for reading a PDF and is steered away from retrying the same failing read.\n\n- Area: Attachments\n- Tier: You'll notice\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}