There are several different methods to upload your files to the platform. Although similar, each upload uses your file slightly differently, which can impact how your model or agent interacts with it. Read about the different uploads to determine which is best suited for your goals and workflow.
Upload to provider
You can upload your files to be processed directly and then used in your conversation. You can use this method for any file, and especially for files that do not contain text content.
- Start a new conversation or open an existing conversation.
- Click the paperclip icon in the text box. A menu opens.
- Select Upload to Provider. Your file directory opens.
- Select a file or files to upload.
- Enter an optional prompt and click enter. Note that the model may take a few seconds to process your file and respond.
Upload as text
You can upload your files as text to allow their full text content to be included directly in your conversation. Note that your model or agent will not actually see that a file was uploaded and instead only be aware of its text content.
- Start a new conversation or open an existing conversation.
- Click the paperclip icon in the text box. A menu opens.
- Select Upload as Text. Your file directory opens.
- Select a file or files to upload.
- Enter an optional prompt and click enter. Note that the model may take a few seconds to process your input and respond.
Upload for file search
File Search capability enables you to upload files for retrieval augmented generation, where model or agent have contextually aware responses based on file subsections. Useful for large documents or multiple files with relevant sections.
- Start a new conversation or open an existing conversation.
- Click the paperclip icon in the text box. A menu opens.
- Select Upload for File Search. Your file directory opens.
- Enter an optional prompt and click enter. Note that the model may take a few seconds to process your file and respond.
