頁面比較

...

Capability/Limitation

Suggested Model	GPT-4 -Turbo-8kseries models
Supported File Types for Upload	Web URL (Supported youtube links)
Supported File Types for OCR	Text in the image from the web URL.
File Restrictions	Web pages that require authentication is not supported, e.g. MTK internal wikis. Web pages with anti-scraping measures is not supported.
Confidential Doc/Data Upload	Please refer to the PIM-related explanations in the FAQ section.
Support for Generating Suggestions	Yes

...

Step1：Click on "+ Chat" to start a new conversation
Image RemovedImage Added
Step2: Click "+ URL" to add the web address you want to analyze (any website, PDF link, or YouTube URL)
Image Added
Image RemovedImage Removed
Image Added

Step3: Select the appropriate Model and Conversation Style based on your needs
Image RemovedImage Added
Step4: After selecting the file(s) you wish to analyze (single or multiple selection), enter your prompt, then you can start asking DaVinci all the answers to the questions you want to know!
Image RemovedImage Added

...

從現在起，你可將你所感興趣的網站、PDF 連結以及 YouTube 影片網址上傳至 MediaTek DaVinci，就能立即向 MediaTek DaVinci 提出問題或請他進行總結，並支持多個連結同時解析。

...

面板

上傳 URL 後，請記得將要使用的 URL 做勾選，可以勾選單一或多 URL

...

建議使用模型	GPT-4 -Turbo-8kseries models
支援上傳的檔案類型	Web URL (包含 youtube)
支援 OCR 的檔案類型	Web URL 中的圖片文字
檔案限制	不支援需要認證的網頁 e.g. 公司內部 wiki 不支援反爬蟲的網頁
機密資訊/檔案上傳說明	請參閱 FAQ 中 PIM 相關說明
支援產生 Suggestion	Yes

...

Step2：點擊右上角的 + URL 新增想分析的網址 (任意網站、PDF 連結或者 YouTube 網址連結)
Image RemovedImage RemovedImage Added
Image Added
Step3：根據需求選擇適合的 Model 和 Conversation Style
Image RemovedImage Added

Step4：勾選後想分析的檔案(可單選或多選)並輸入你的 Prompt，即可開始向 MediaTek DaVinci 詢問所有你想知道問題的答案！
Image RemovedImage Added

...

...