Getting whole-page images instead of embedded figures - use JSON mode

TL;DR

Check the JSON image objects first; that is where the extracted figures and their descriptions are. A working solution was posted in the thread and confirmed by another user.

Fix

If LlamaParse gives you whole-page images instead of the embedded figures, do not fight the default markdown output - use JSON mode and dig into the image objects in the returned JSON, which is where the per-image results live. In markdown mode you can also get image descriptions by stating in the parsing instructions which images to describe. Check the JSON image objects first; that is where the extracted figures and their descriptions are.

Context: Issue run-llama/llamacloudservices#317 (closed, 9 comments): a user calling getjsonresult() wanted the individual embedded images from each PDF page, but kept getting one whole-page image per page instead. A working solution was posted in the thread and confirmed by another user.

Matched source

Source: Published skill Original query: "Getting whole-page images instead of embedded figures - use JSON mode" Key terms: embedded, figures, getting, images, instead, json, mode, page, whole

When to use this

• Use this when your task, error, or question matches the title above and the setup in the Fix section.

When not to use this

• Skip it when the error text, tool, or situation is different. Forcing a mismatched fix usually costs more time than searching again.

Compatibility

The original write-up does not pin tool versions. Treat the commands and UI paths above as reflecting the author's setup, and check your tool's own docs if your version looks different.

Source: Issue run-llama/llamacloudservices#317 (closed, 9 comments): a user calling getjsonresult() wanted the individual embedded images from each PDF page, but kept getting one whole-page image per page instead. A working solution was posted in the thread and confirmed by another user.

Verified resolution: If LlamaParse gives you whole-page images instead of the embedded figures, do not fight the default markdown output - use JSON mode and dig into the image objects in the returned JSON, which is where the per-image results live. In markdown mode you can also get image descriptions by stating in the parsing instructions which images to describe. Check the JSON image objects first; that is where the extracted figures and their descriptions are.

Source URL: https://github.com/run-llama/llamacloudservices/issues/317