From 9d15453b668685e1b85694d7af439691cf67b5cf Mon Sep 17 00:00:00 2001 From: new5558 Date: Fri, 22 Aug 2025 09:39:34 +0700 Subject: [PATCH] docs: update jan v1 research blog title and keywords 3 --- docs/src/pages/post/jan-v1-for-research.mdx | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/src/pages/post/jan-v1-for-research.mdx b/docs/src/pages/post/jan-v1-for-research.mdx index 73a622401..6b503f718 100644 --- a/docs/src/pages/post/jan-v1-for-research.mdx +++ b/docs/src/pages/post/jan-v1-for-research.mdx @@ -36,7 +36,7 @@ This guide explores how to optimize Jan-V1 for research-intensive tasks by switc Jan-V1 comes with a built-in search and scrape prompt in its default template. This prompt is optimized for SimpleQA-style questions and daily tasks that require short, accurate responses. However, our testing revealed that when tasked with generating comprehensive long-form reports or conducting in-depth research, the model's performance was inconsistent with the default prompt. -We experimented with Jan-V1 using a new system prompt optimized for research-intensive tasks and found significant improvements in report generation capability based on our internal benchmarks. The research prompt produces more comprehensive, well-structured reports with proper citations and deeper analysis compared to the default prompt. +We experimented with Jan-V1 using a new system prompt optimized for research-intensive tasks and found significant improvements in report generation capability. The research prompt produces more comprehensive, well-structured reports with proper citations and deeper analysis compared to the default prompt. ---