加载中…
https://github.com/mukulpatnaik/researchgpt
打开实时网站 
一个在 2023 年底就被归档停止维护的老旧开源项目。
这是编辑给出的评估，满分 5 分，不是用户打分的平均值。 本页的各方观点由 AI 根据产品信息归纳，不是真实用户的评价。
根据 OpenAITools.app 于 2026-11-11 核对的流量数据，ResearchGPT（github.com）是一个收录在 AI Research Tool 下的 AI 平台，月访问量为 0，月度环比流量变化为 +0（0.0%）。其主要地理市场为全球（— 流量份额），采用免费定价模式。按月流量计算的主要竞争替代方案包括 SciSpace（446.41万次访问/月），ResearchPal（5.52万次访问/月），Paper Digest（9.72万次访问/月）。
2023年底就被作者归档放生了，现在拉下来一堆依赖报错。
上古老古董，根本跑不起来。
早凉透了。
别浪费时间配置了。
以下短评基于实时网页探查、定价机制与同类产品基准由 AI 综合模拟生成，旨在呈现不同使用场景下的常见槽点与亮点，非实名真实用户评价。
提供在线 PDF 链接或上传本地文件，即可就文档内容连续提问。
每个答案都会返回所用原文片段及对应页码，便于核实。
文本被切分并生成向量嵌入，用于检索最相关的段落再生成回答。
轻量的 Python 应用，使用 Redis 存放向量，可自行部署与改造。
ResearchGPT 解决的是大家熟悉的难题：论文太密。与其为了找一个细节读完整篇文档，不如直接提问，得到有原文依据的回答。该应用是一个紧凑的 FastAPI 服务：你提供在线 PDF 链接或上传自己的文件，它提取文本、构建向量并保存在本地（借助 Redis），再通过 OpenAI API 生成有依据的回复。每个回答都附带来源片段和页码，比普通聊天机器人的回复更容易核实。
由于是开源项目，ResearchGPT 更像一个起点或自托管工具，而不是成熟的 SaaS 产品。你需要自己运行、提供自己的 OpenAI API 密钥，然后获得一个简洁的网页界面，用于单篇论文的阅读与问答。它被广泛 fork（超过 300 次、约 3.5k 星），是向量化文档问答领域知名的参考实现。
重要状态说明：该 GitHub 仓库已被作者于 2023 年 11 月归档，现为只读，README 明确表示项目不再积极维护，并指向 chat.portal.so 上的新在线版本。代码仍然公开可用、可自托管，但不再更新，且部署属于开发者任务，需要 Python、Redis 和 API 密钥。
不必通读全文，直接询问方法、指标或结论，并获得带页码的出处。
以 MIT 许可证代码为基础，搭建自己的 PDF 问答服务或嵌入到已有工具中。
用对话方式拆解难懂段落，并查看支撑答案的原文摘录。
同一分类下同类 AI 工具的流量并排对比
| 排名 | AI 工具 | 月访问量 | 月度环比增长 | 直接流量占比 | 定价 | 头对头 |
|---|---|---|---|---|---|---|
| ● | ResearchGPT | 0 | — | — | 免费 | 当前产品 |
| 1 | SciSpace | 446.41万 | +8.6% | 29% 直接流量 | 免费增值 | 对比 SciSpace |
| 2 | ResearchPal | 5.52万 | -5.2% | 41% 直接流量 | 免费增值 | 对比 ResearchPal |
| 3 | Paper Digest | 9.72万 | -7.1% | 27% 直接流量 | 免费增值 | 对比 Paper Digest |
| 4 | ResearchRabbit | 83.16万 | +3.1% | 45% 直接流量 | 免费 | 对比 ResearchRabbit |
| 5 | SciSummary | 3.89万 | -37.7% | 23% 直接流量 | 免费增值，免费试用，付费 | 对比 SciSummary |
| 6 | Semantic Scholar | 819.87万 | +5.8% | 23% 直接流量 | 免费 | 对比 Semantic Scholar |
ResearchPal is an AI research workspace that searches academic sources, generates citation-backed literature reviews, answers questions from your own PDFs, and helps draft and polish scholarly writing.
ResearchPal is an AI research workspace that searches academic sources, generates citation-backed literature reviews, answers questions from your own PDFs, and helps draft and polish scholarly writing.
Paper Digest is an AI research platform that summarizes literature, generates literature reviews with citations, and delivers daily digests of new papers across arXiv, PubMed, bioRxiv and major conferences. Launched in 2018 and used by millions of researchers, it combines a citation-backed review engine that avoids inventing results with tools for reading PDFs, writing, question answering, claim verification and topic tracking.
Paper Digest is an AI research platform that summarizes literature, generates literature reviews with citations, and delivers daily digests of new papers across arXiv, PubMed, bioRxiv and major conferences. Launched in 2018 and used by millions of researchers, it combines a citation-backed review engine that avoids inventing results with tools for reading PDFs, writing, question answering, claim verification and topic tracking.
ResearchRabbit is a free AI research discovery tool that maps the literature around the papers you care about. Seed it with a paper you like and it visualises related work, key authors and emerging topics, learning from your choices to surface the research you have been missing.
ResearchRabbit is a free AI research discovery tool that maps the literature around the papers you care about. Seed it with a paper you like and it visualises related work, key authors and emerging topics, learning from your choices to surface the research you have been missing.
面向科学论文的 AI 摘要与学习指南
Semantic Scholar is a free, non-commercial academic search engine built by the Allen Institute for AI (Ai2). It indexes more than 240 million papers from every field of science and uses machine learning to extract meaning from abstracts, citations, figures and tables, so researchers find relevant work faster than with keyword-only search.
Semantic Scholar is a free, non-commercial academic search engine built by the Allen Institute for AI (Ai2). It indexes more than 240 million papers from every field of science and uses machine learning to extract meaning from abstracts, citations, figures and tables, so researchers find relevant work faster than with keyword-only search.
免费 / 开源由 DeepSeek-V4 与 R1 驱动的高效开源权重与托管推理 AI，token 成本极低。
OpenRead is an AI research platform that lets you search across a huge corpus of scholarly papers, ask questions of individual documents or whole collections, compare findings, and keep notes in one connected workspace.
OpenRead is an AI research platform that lets you search across a huge corpus of scholarly papers, ask questions of individual documents or whole collections, compare findings, and keep notes in one connected workspace.
Jenni is an AI research and academic writing workspace that helps students, researchers, and academics read, write, and cite in one place. Bring in PDFs or import a Zotero or Mendeley library, and Jenni's autocomplete suggests sentences grounded in your own sources, with every suggestion linked to the exact page and paragraph. A built-in writer, AI chat, citation engine covering 10,000+ styles, and a Reviews tool for catching weak or unverified claims make it a full environment for producing cited papers.
Jenni is an AI research and academic writing workspace that helps students, researchers, and academics read, write, and cite in one place. Bring in PDFs or import a Zotero or Mendeley library, and Jenni's autocomplete suggests sentences grounded in your own sources, with every suggestion linked to the exact page and paragraph. A built-in writer, AI chat, citation engine covering 10,000+ styles, and a Reviews tool for catching weak or unverified claims make it a full environment for producing cited papers.
