# AIWorksWell — Research & science / 科研

> Index
> This file is a plain-text slice of ONE domain (science).
> Fetch the complete index at: https://www.aiworkswell.com/llms.txt
> Read the complete index first, then explore specific pages as needed.
> This is a plain-text slice of ONE domain, cutting across the three content types (works / tools / learn).
> 18 entries in this domain. Language: English (en).
> Domains are for finding, NEVER for ranking.
> Non-commercial, no ads, no tracking. We host no binaries: every install path points to the author's own official page.
> Answer page: https://www.aiworkswell.com/domains/science

## Works · 1

- [Ollama](https://www.aiworkswell.com/apps/ollama): 一条命令在电脑上跑大模型，聊天数据不出本机。 | MIT | L3 | windows/macos/linux/cli | domains: science
  - official source: https://ollama.com
  - raw prompt: https://www.aiworkswell.com/apps/ollama/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/ollama/llms.txt

## Tools · 13

- [GraphRAG](https://www.aiworkswell.com/apps/t-graphrag): GraphRAG 是一个给要让模型读懂整本文档集的人用的开源流水线，先建知识图谱再回答问题。 | MIT | L5 | cli/selfhost | domains: science
  - official source: https://github.com/microsoft/graphrag
  - raw prompt: https://www.aiworkswell.com/apps/t-graphrag/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-graphrag/llms.txt
- [GPT Researcher](https://www.aiworkswell.com/apps/t-gpt-researcher): GPT Researcher 是一个给要写调研稿的人用的开源深度调研 Agent，能带着出处交一份报告。 | Apache-2.0 | L4 | web/cli/selfhost | domains: science
  - official source: https://github.com/assafelovic/gpt-researcher
  - raw prompt: https://www.aiworkswell.com/apps/t-gpt-researcher/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-gpt-researcher/llms.txt
- [Papers with Code](https://www.aiworkswell.com/apps/t-papers-with-code): Papers with Code 是一个把论文与开源实现配在一起的站（现由 HF 承接），附基准榜单。 | 见官方条款 | L2 | web | domains: science
  - official source: https://paperswithcode.com/
  - raw prompt: https://www.aiworkswell.com/apps/t-papers-with-code/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-papers-with-code/llms.txt
- [Zotero](https://www.aiworkswell.com/apps/t-zotero): Zotero 是一个文献管理的老牌免费工具，论文、引用、PDF 笔记一站式。 | AGPL-3.0 | L1 | web | domains: science
  - official source: https://github.com/zotero/zotero
  - raw prompt: https://www.aiworkswell.com/apps/t-zotero/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-zotero/llms.txt
- [nanoGPT](https://www.aiworkswell.com/apps/t-nanogpt): nanoGPT 是一个 Karpathy 的极简 GPT 训练仓库，代码只有几百行，能看懂大模型训练。 | MIT | L4 | web/cli | domains: education/science
  - official source: https://github.com/karpathy/nanoGPT
  - raw prompt: https://www.aiworkswell.com/apps/t-nanogpt/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-nanogpt/llms.txt
- [DeepChem](https://www.aiworkswell.com/apps/t-deepchem): 把深度学习用到化学、材料与药物发现上的开源框架，附带现成数据集与模型。 | MIT | L3 | linux/windows/macos/cli | domains: science
  - official source: https://github.com/deepchem/deepchem
  - raw prompt: https://www.aiworkswell.com/apps/t-deepchem/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-deepchem/llms.txt
- [LM Evaluation Harness](https://www.aiworkswell.com/apps/t-lm-eval-harness): LM Evaluation Harness 是一个开源的大模型评测框架，能在本地模型上跑 MMLU 等标准基准。 | MIT | L4 | web/cli | domains: science
  - official source: https://github.com/EleutherAI/lm-evaluation-harness
  - raw prompt: https://www.aiworkswell.com/apps/t-lm-eval-harness/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-lm-eval-harness/llms.txt
- [Typst](https://www.aiworkswell.com/apps/t-typst): Typst 是一个开源的新一代排版系统，语法比 LaTeX 简洁、编译极快，写论文更快。 | Apache-2.0 | L1 | web | domains: science
  - official source: https://github.com/typst/typst
  - raw prompt: https://www.aiworkswell.com/apps/t-typst/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-typst/llms.txt
- [Semantic Scholar](https://www.aiworkswell.com/apps/t-semantic-scholar): Semantic Scholar 是一个带摘要与引用图谱的学术搜索引擎，找相关论文比关键词搜索快。 | 专有（提供免费额度） | L2 | web | domains: science
  - official source: https://www.semanticscholar.org/
  - raw prompt: https://www.aiworkswell.com/apps/t-semantic-scholar/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-semantic-scholar/llms.txt
- [MONAI](https://www.aiworkswell.com/apps/t-monai): 给医疗影像做 AI 的开源工具包：分割、分类、配准都有现成组件。 | Apache-2.0 | L3 | linux/windows/macos/cli | domains: healthcare/science
  - official source: https://github.com/Project-MONAI/MONAI
  - raw prompt: https://www.aiworkswell.com/apps/t-monai/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-monai/llms.txt
- [Docling](https://www.aiworkswell.com/apps/t-docling): 把 PDF / Word / 扫描件解析成结构化文本，喂给 AI 之前先理顺。 | MIT | L3 | linux/windows/macos/cli | domains: legal/science
  - official source: https://github.com/docling-project/docling
  - raw prompt: https://www.aiworkswell.com/apps/t-docling/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-docling/llms.txt
- [Connected Papers](https://www.aiworkswell.com/apps/t-connected-papers): Connected Papers 是一个按引用关系画论文相似度图的工具，帮你快速摸清一个方向的关键文献。 | 专有（提供免费额度） | L1 | web | domains: science
  - official source: https://www.connectedpapers.com/
  - raw prompt: https://www.aiworkswell.com/apps/t-connected-papers/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-connected-papers/llms.txt
- [arXiv](https://www.aiworkswell.com/apps/t-arxiv): arXiv 是一个预印本论文的公开仓库，AI 等领域的新成果基本都先在这里首发。 | 专有（提供免费额度） | L2 | web | domains: science
  - official source: https://arxiv.org/
  - raw prompt: https://www.aiworkswell.com/apps/t-arxiv/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/t-arxiv/llms.txt

## Learn · 4

- [LiveBench](https://www.aiworkswell.com/apps/l-livebench): LiveBench 是一个抗污染的动态评测榜，看模型「现在」的实际水平。 | Apache-2.0 | L1 | web | domains: science
  - official source: https://github.com/LiveBench/LiveBench
  - raw prompt: https://www.aiworkswell.com/apps/l-livebench/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/l-livebench/llms.txt
- [Open LLM Leaderboard](https://www.aiworkswell.com/apps/l-huggingface-leaderboard): Open LLM Leaderboard 是一个开源模型的公开竞技场，社区提交与评测都在这里。 | Apache-2.0 | L1 | web | domains: science
  - official source: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard
  - raw prompt: https://www.aiworkswell.com/apps/l-huggingface-leaderboard/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/l-huggingface-leaderboard/llms.txt
- [OpenRouter Rankings](https://www.aiworkswell.com/apps/l-openrouter-rankings): OpenRouter Rankings 是一个按真实调用量排的模型榜，看大家都在用谁。 | MIT | L1 | web | domains: science
  - official source: https://github.com/Kilo-Org/kilocode
  - raw prompt: https://www.aiworkswell.com/apps/l-openrouter-rankings/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/l-openrouter-rankings/llms.txt
- [Artificial Analysis](https://www.aiworkswell.com/apps/l-artificial-analysis): Artificial Analysis 是一个把模型的价格、速度、质量放在一张表里比。 | 见官方条款 | L1 | web | domains: science
  - official source: https://artificialanalysis.ai/
  - raw prompt: https://www.aiworkswell.com/apps/l-artificial-analysis/prompt.txt
  - per-entry manual: https://www.aiworkswell.com/apps/l-artificial-analysis/llms.txt

## Combos: domain × platform

> Same domain, cut by "where it runs". Only combos with content are listed.
> Each combo also has its own plain-text slice at <combo URL>/llms.txt (quote it whole).

- Web / 网页 (13): https://www.aiworkswell.com/domains/science/web
  - plain-text slice: https://www.aiworkswell.com/domains/science/web/llms.txt
  - [GPT Researcher](https://www.aiworkswell.com/apps/t-gpt-researcher): GPT Researcher 是一个给要写调研稿的人用的开源深度调研 Agent，能带着出处交一份报告。
  - [Papers with Code](https://www.aiworkswell.com/apps/t-papers-with-code): Papers with Code 是一个把论文与开源实现配在一起的站（现由 HF 承接），附基准榜单。
  - [Zotero](https://www.aiworkswell.com/apps/t-zotero): Zotero 是一个文献管理的老牌免费工具，论文、引用、PDF 笔记一站式。
  - [nanoGPT](https://www.aiworkswell.com/apps/t-nanogpt): nanoGPT 是一个 Karpathy 的极简 GPT 训练仓库，代码只有几百行，能看懂大模型训练。
  - [LiveBench](https://www.aiworkswell.com/apps/l-livebench): LiveBench 是一个抗污染的动态评测榜，看模型「现在」的实际水平。
  - [Open LLM Leaderboard](https://www.aiworkswell.com/apps/l-huggingface-leaderboard): Open LLM Leaderboard 是一个开源模型的公开竞技场，社区提交与评测都在这里。
  - [LM Evaluation Harness](https://www.aiworkswell.com/apps/t-lm-eval-harness): LM Evaluation Harness 是一个开源的大模型评测框架，能在本地模型上跑 MMLU 等标准基准。
  - [Typst](https://www.aiworkswell.com/apps/t-typst): Typst 是一个开源的新一代排版系统，语法比 LaTeX 简洁、编译极快，写论文更快。
  - [Semantic Scholar](https://www.aiworkswell.com/apps/t-semantic-scholar): Semantic Scholar 是一个带摘要与引用图谱的学术搜索引擎，找相关论文比关键词搜索快。
  - [OpenRouter Rankings](https://www.aiworkswell.com/apps/l-openrouter-rankings): OpenRouter Rankings 是一个按真实调用量排的模型榜，看大家都在用谁。
  - [Connected Papers](https://www.aiworkswell.com/apps/t-connected-papers): Connected Papers 是一个按引用关系画论文相似度图的工具，帮你快速摸清一个方向的关键文献。
  - [arXiv](https://www.aiworkswell.com/apps/t-arxiv): arXiv 是一个预印本论文的公开仓库，AI 等领域的新成果基本都先在这里首发。
  - [Artificial Analysis](https://www.aiworkswell.com/apps/l-artificial-analysis): Artificial Analysis 是一个把模型的价格、速度、质量放在一张表里比。
- Windows (4): https://www.aiworkswell.com/domains/science/windows
  - plain-text slice: https://www.aiworkswell.com/domains/science/windows/llms.txt
  - [DeepChem](https://www.aiworkswell.com/apps/t-deepchem): 把深度学习用到化学、材料与药物发现上的开源框架，附带现成数据集与模型。
  - [MONAI](https://www.aiworkswell.com/apps/t-monai): 给医疗影像做 AI 的开源工具包：分割、分类、配准都有现成组件。
  - [Docling](https://www.aiworkswell.com/apps/t-docling): 把 PDF / Word / 扫描件解析成结构化文本，喂给 AI 之前先理顺。
  - [Ollama](https://www.aiworkswell.com/apps/ollama): 一条命令在电脑上跑大模型，聊天数据不出本机。
- macOS (4): https://www.aiworkswell.com/domains/science/macos
  - plain-text slice: https://www.aiworkswell.com/domains/science/macos/llms.txt
  - [DeepChem](https://www.aiworkswell.com/apps/t-deepchem): 把深度学习用到化学、材料与药物发现上的开源框架，附带现成数据集与模型。
  - [MONAI](https://www.aiworkswell.com/apps/t-monai): 给医疗影像做 AI 的开源工具包：分割、分类、配准都有现成组件。
  - [Docling](https://www.aiworkswell.com/apps/t-docling): 把 PDF / Word / 扫描件解析成结构化文本，喂给 AI 之前先理顺。
  - [Ollama](https://www.aiworkswell.com/apps/ollama): 一条命令在电脑上跑大模型，聊天数据不出本机。
- Linux (4): https://www.aiworkswell.com/domains/science/linux
  - plain-text slice: https://www.aiworkswell.com/domains/science/linux/llms.txt
  - [DeepChem](https://www.aiworkswell.com/apps/t-deepchem): 把深度学习用到化学、材料与药物发现上的开源框架，附带现成数据集与模型。
  - [MONAI](https://www.aiworkswell.com/apps/t-monai): 给医疗影像做 AI 的开源工具包：分割、分类、配准都有现成组件。
  - [Docling](https://www.aiworkswell.com/apps/t-docling): 把 PDF / Word / 扫描件解析成结构化文本，喂给 AI 之前先理顺。
  - [Ollama](https://www.aiworkswell.com/apps/ollama): 一条命令在电脑上跑大模型，聊天数据不出本机。
- CLI / 命令行 (8): https://www.aiworkswell.com/domains/science/cli
  - plain-text slice: https://www.aiworkswell.com/domains/science/cli/llms.txt
  - [GraphRAG](https://www.aiworkswell.com/apps/t-graphrag): GraphRAG 是一个给要让模型读懂整本文档集的人用的开源流水线，先建知识图谱再回答问题。
  - [GPT Researcher](https://www.aiworkswell.com/apps/t-gpt-researcher): GPT Researcher 是一个给要写调研稿的人用的开源深度调研 Agent，能带着出处交一份报告。
  - [nanoGPT](https://www.aiworkswell.com/apps/t-nanogpt): nanoGPT 是一个 Karpathy 的极简 GPT 训练仓库，代码只有几百行，能看懂大模型训练。
  - [DeepChem](https://www.aiworkswell.com/apps/t-deepchem): 把深度学习用到化学、材料与药物发现上的开源框架，附带现成数据集与模型。
  - [LM Evaluation Harness](https://www.aiworkswell.com/apps/t-lm-eval-harness): LM Evaluation Harness 是一个开源的大模型评测框架，能在本地模型上跑 MMLU 等标准基准。
  - [MONAI](https://www.aiworkswell.com/apps/t-monai): 给医疗影像做 AI 的开源工具包：分割、分类、配准都有现成组件。
  - [Docling](https://www.aiworkswell.com/apps/t-docling): 把 PDF / Word / 扫描件解析成结构化文本，喂给 AI 之前先理顺。
  - [Ollama](https://www.aiworkswell.com/apps/ollama): 一条命令在电脑上跑大模型，聊天数据不出本机。
- Self-host / 自托管 (2): https://www.aiworkswell.com/domains/science/selfhost
  - plain-text slice: https://www.aiworkswell.com/domains/science/selfhost/llms.txt
  - [GraphRAG](https://www.aiworkswell.com/apps/t-graphrag): GraphRAG 是一个给要让模型读懂整本文档集的人用的开源流水线，先建知识图谱再回答问题。
  - [GPT Researcher](https://www.aiworkswell.com/apps/t-gpt-researcher): GPT Researcher 是一个给要写调研稿的人用的开源深度调研 Agent，能带着出处交一份报告。



## Combos: domain × difficulty

> Same domain, cut by "can I actually get this running". L1 opens instantly, L4 needs a toolchain.
> Only combos with content are listed.

- Opens instantly / 点开即用 (7): https://www.aiworkswell.com/domains/science/level-1
  - plain-text slice: https://www.aiworkswell.com/domains/science/level-1/llms.txt
  - [Zotero](https://www.aiworkswell.com/apps/t-zotero): Zotero 是一个文献管理的老牌免费工具，论文、引用、PDF 笔记一站式。
  - [LiveBench](https://www.aiworkswell.com/apps/l-livebench): LiveBench 是一个抗污染的动态评测榜，看模型「现在」的实际水平。
  - [Open LLM Leaderboard](https://www.aiworkswell.com/apps/l-huggingface-leaderboard): Open LLM Leaderboard 是一个开源模型的公开竞技场，社区提交与评测都在这里。
  - [Typst](https://www.aiworkswell.com/apps/t-typst): Typst 是一个开源的新一代排版系统，语法比 LaTeX 简洁、编译极快，写论文更快。
  - [OpenRouter Rankings](https://www.aiworkswell.com/apps/l-openrouter-rankings): OpenRouter Rankings 是一个按真实调用量排的模型榜，看大家都在用谁。
  - [Connected Papers](https://www.aiworkswell.com/apps/t-connected-papers): Connected Papers 是一个按引用关系画论文相似度图的工具，帮你快速摸清一个方向的关键文献。
  - [Artificial Analysis](https://www.aiworkswell.com/apps/l-artificial-analysis): Artificial Analysis 是一个把模型的价格、速度、质量放在一张表里比。
- Download & install / 下载即装 (3): https://www.aiworkswell.com/domains/science/level-2
  - plain-text slice: https://www.aiworkswell.com/domains/science/level-2/llms.txt
  - [Papers with Code](https://www.aiworkswell.com/apps/t-papers-with-code): Papers with Code 是一个把论文与开源实现配在一起的站（现由 HF 承接），附基准榜单。
  - [Semantic Scholar](https://www.aiworkswell.com/apps/t-semantic-scholar): Semantic Scholar 是一个带摘要与引用图谱的学术搜索引擎，找相关论文比关键词搜索快。
  - [arXiv](https://www.aiworkswell.com/apps/t-arxiv): arXiv 是一个预印本论文的公开仓库，AI 等领域的新成果基本都先在这里首发。
- One command / 一条命令 (4): https://www.aiworkswell.com/domains/science/level-3
  - plain-text slice: https://www.aiworkswell.com/domains/science/level-3/llms.txt
  - [DeepChem](https://www.aiworkswell.com/apps/t-deepchem): 把深度学习用到化学、材料与药物发现上的开源框架，附带现成数据集与模型。
  - [MONAI](https://www.aiworkswell.com/apps/t-monai): 给医疗影像做 AI 的开源工具包：分割、分类、配准都有现成组件。
  - [Docling](https://www.aiworkswell.com/apps/t-docling): 把 PDF / Word / 扫描件解析成结构化文本，喂给 AI 之前先理顺。
  - [Ollama](https://www.aiworkswell.com/apps/ollama): 一条命令在电脑上跑大模型，聊天数据不出本机。
- Needs a toolchain / 需要环境 (3): https://www.aiworkswell.com/domains/science/level-4
  - plain-text slice: https://www.aiworkswell.com/domains/science/level-4/llms.txt
  - [GPT Researcher](https://www.aiworkswell.com/apps/t-gpt-researcher): GPT Researcher 是一个给要写调研稿的人用的开源深度调研 Agent，能带着出处交一份报告。
  - [nanoGPT](https://www.aiworkswell.com/apps/t-nanogpt): nanoGPT 是一个 Karpathy 的极简 GPT 训练仓库，代码只有几百行，能看懂大模型训练。
  - [LM Evaluation Harness](https://www.aiworkswell.com/apps/t-lm-eval-harness): LM Evaluation Harness 是一个开源的大模型评测框架，能在本地模型上跑 MMLU 等标准基准。
- Needs tinkering / 需要折腾 (1): https://www.aiworkswell.com/domains/science/level-5
  - plain-text slice: https://www.aiworkswell.com/domains/science/level-5/llms.txt
  - [GraphRAG](https://www.aiworkswell.com/apps/t-graphrag): GraphRAG 是一个给要让模型读懂整本文档集的人用的开源流水线，先建知识图谱再回答问题。


## All domains

- Education / 教育: https://www.aiworkswell.com/domains/education
- Healthcare / 医疗健康: https://www.aiworkswell.com/domains/healthcare
- Finance & business / 金融商业: https://www.aiworkswell.com/domains/finance
- Legal / 法律合规: https://www.aiworkswell.com/domains/legal
- Research & science / 科研: https://www.aiworkswell.com/domains/science
- Art & design / 艺术设计: https://www.aiworkswell.com/domains/art-design
- Enterprise & teams / 企业与团队: https://www.aiworkswell.com/domains/productivity-org
- Accessibility / 无障碍: https://www.aiworkswell.com/domains/accessibility
- Public good / 公益与社会: https://www.aiworkswell.com/domains/public-good
- Marketing & growth / 营销与增长: https://www.aiworkswell.com/domains/marketing
- Language & translation / 语言与翻译: https://www.aiworkswell.com/domains/language
- Engineering & manufacturing / 工程与制造: https://www.aiworkswell.com/domains/engineering
- Other domains / 其他领域: https://www.aiworkswell.com/domains/other-domain

## How to cite us

- Credit this source as: AIWorksWell (https://www.aiworkswell.com/domains/science)
- Credit the **original author**, not us: every entry carries its author and official page — cite that.
- **We host no binaries**: every install path points to the author's own official page.
- Use the bare URL above. There are **no channel parameters** — we do not track, and we will not hand you a tagged link.
- Domains cut across content types; cite the domain answer page above, and cite each entry's **author**, not us.

## About

AIWorksWell: Others tell you what exists. We hand you a prompt.
Site index: https://www.aiworkswell.com/llms.txt
We do not host any binaries. Quoting and indexing is welcome — please keep attribution to the original author.