Parse

Question

Hello everyone,I came across the "user_prompt," "system_prompt," and "system_prompt_append" parameters in the LlamaParse documentation, but I wasn’t able to use them successfully through the API. I couldn't find them in the implementation, and my attempts to use them didn’t work as expected.Given this, do we still need to use "parsing_instruction" until the next update, or is there another approach to achieve the same functionality?I would greatly appreciate any insights or clarification on this. Thank you in advance.

Logan M · Answer

The latest version of the sdk should support it? pip install -U llama-parse ?But yes in the meantime parsing_instruction works fine

Toshi · Answer

@Logan M I ran the command you suggested (pip install -U llama-parse), and it returned multiple lines with "Requirement already satisfied:", confirming that I already have the latest version installed (v0.6.1).However, I still couldn't get the "user_prompt", "system_prompt", and "system_prompt_append" parameters to work. In my last test, I used the LlamaParse method as follows:parser = LlamaParse( api_key=os.getenv("API_KEY"), result_type=ResultType.MD, use_vendor_multimodal_model=True, vendor_multimodal_model_name="gemini-2.0-flash-001", language="pt", do_not_cache=True, parsing_mode="parse_page_with_lvm", user_prompt=[text], system_prompt_append=[text], system_prompt=[text]
)
file_path = [file_path]
documents = await parser.aload_data(file_path)Unfortunately, this implementation only returned a simple OCR result instead of the expected response. (job id: 723f549a-b939-4762-ab94-1efc95f32f20)Is there any additional parameter that needs to be set, or am I using these parameters incorrectly?I would greatly appreciate any insights or guidance you could provide. Thank you in advance.

Find answers from the community

Parse