AI Tools for Science and Research

Signup

Go to: All AI Tools AI Tools by Tool Type AI Tools by Job Type

AI Tools for Science and Research

Accelerate your scientific discoveries and research projects using advanced AI tools for data analysis, simulation, visualization, and automated experimentation, enhancing accuracy and efficiency.

Signup & Stay Updated on the latest AI Tools

AI Tools for Science and Research

### AI Tools available

Keysake

Keysake helps product teams and UX researchers capture, organize, and revisit interview recordings and user research. It is a research companion that generates plain-language recaps and lets you search across your sessions.

Sider Omni Sidebar

Sider Omni Sidebar is an AI assistant that integrates into your browser sidebar to summarize web pages and chat with PDFs or images. It is built for professionals and students who want to handle research and reading tasks without switching tabs.

S-Roll

S-Roll is an agentic harness that finds and clips moments in long videos using natural-language prompts. It runs locally on Apple silicon and is for video creators, editors, and researchers who want to search recordings without uploading them.

Compute:Arena

Compute:Arena is an open-source benchmarking tool for testing how fast local language models run on specific hardware. It lets anyone measure performance and submit results to a public leaderboard, helping users and developers compare model speeds...

LM-Kit One

LM-Kit One is a private AI application server that combines model inference, knowledge search, agents, and orchestration through a REST API. It runs on your own infrastructure using your chosen models and is free for personal, educational, and eli...

Codex GPU Queue

Codex GPU Queue manages and optimizes access to shared GPU resources for development teams. It lets users submit, prioritize, and track compute jobs across available hardware.

Web Search Agents by Nimble

Web Search Agents by Nimble are specialized web experts that handle complex research, data enrichment, and dataset creation tasks. They are built for developers and technical founders who need to feed large language models with relevant, high-qual...

AlphaGenome Atlas

AlphaGenome Atlas lets researchers explore predicted effects of over 9 billion single-letter DNA mutations. It is built for biologists and geneticists who need to query variant impacts without writing code.

DuckFightClub

DuckFightClub is a community and event platform where robot enthusiasts train virtual Microduck bipeds in sumo wrestling using a shared simulator. Participants compete to see whose trained model wins, with events held online and in person.

Catenary

Catenary is a spatial IDE and terminal orchestrator for developers working with AI coding agents. It provides an infinite 2D canvas to visually connect terminals, editors, and browsers, allowing you to chain models and manage isolated workspaces. ...

Loqua

Loqua is a voice assistant built on an omni model that understands both speech and what's on your screen. It is for researchers and professionals who want to speak naturally to ask about documents or code, capture context, and delegate multi-step ...

Clipnote

Clipnote lets you save, organize, and share AI-generated content such as Markdown or HTML. It is for anyone using tools like ChatGPT, Claude, or Gemini who needs a central place to keep their outputs. Content can be added directly from an AI chat ...

WeatherNext 3

WeatherNext 3 provides weather forecasts that include solar irradiance and wind speed variables. It is built for grid operators and energy traders who need to predict renewable generation output based on weather conditions.

Readr

Readr is a document reader for DRM-free EPUB, PDF, and Markdown files that lets you ask questions about selected passages. It is for readers who want to query their own books using their own AI keys and get cited, book-sourced answers without an a...

Causal

Causal is an infinite canvas for creatives to store, connect, and plan visual ideas. Its integrated AI understands spatial relationships across files, notes, and images, and can generate and arrange content directly on the board.

Claude Fable 5.1

Claude Fable 5.1 is an AI model for developers building long-running, agentic workflows. It reduces typical costs by 25 percent and cuts cybersecurity false positives by 60 percent.

TrustedRouter

TrustedRouter lets you use many AI models without sending your data to a third party. It is built for users and developers who want private, direct access to LLMs and tools for comparing model performance.

Murfy AI

Murfy AI is an AI agent for researchers and technical writers that assists with drafting papers, reviewing manuscripts, revising based on reviewer comments, fixing LaTeX compile errors, and building Beamer slides. It is designed for academics and ...

Hy4 preview

Hy4 preview is a large language model from Tencent designed for long-horizon agentic tasks. It autonomously handles coding, game development, and complex document analysis, running tests and fixing bugs before delivering results. It is intended fo...

Zero

Zero is an AI tool that generates concise summaries of long documents, articles, and reports. It is intended for professionals and students who need to quickly extract key information without reading the full text. The tool outputs plain-language ...

Project SKY

Project SKY is an ambient desktop companion for Windows that provides real-time screen understanding, voice conversation, and persistent memory across sessions. It is built for developers, designers, and researchers who want AI assistance without ...

KerasFormers

KerasFormers is a library for running Hugging Face model architectures in Keras 3, supporting over 60 vision, speech, large language, and vision-language models. It provides weight conversion and parity checks so users can work across TensorFlow, ...

Local

Local is an AI tool that runs models directly on your device, eliminating the need for cloud servers. It automatically configures itself for your hardware and recommends models that your system can run. It is designed for people who want private, ...

Glasp for Firefox

Glasp for Firefox is a browser extension for highlighting text, images, PDFs, and YouTube transcripts on web pages. It also supports adding notes and tags, summarizing content with AI chatbots, and syncing all saved items to a searchable library t...

Ressearch AI

Ressearch AI connects research questions, literature, analysis, and writing into one workflow for life scientists and graduate students. It uses specialized agents to search scientific sources, run reproducible Python or R analyses in cloud sandbo...

Skim Recap

Skim Recap generates summaries of web articles you scroll past quickly. It is designed for readers who want to review skipped content without rereading the entire page. The tool runs locally in your browser using WebGPU.

Zetik

Zetik is an AI tool that scans early-stage companies and projects to identify the few developments most relevant to a user's specified interests. It is designed for professionals who need to monitor large volumes of emerging information, sorting f...

ThreadPort

ThreadPort is a browser extension that transfers an active AI chat between tools like ChatGPT, Claude, and Gemini with one click, preserving formatting and context. It is built for users who regularly switch AI assistants and want to move current ...

nenspace

nenspace combines the nen family of language models-trained to question assumptions instead of agreeing with users-with a workspace for tasks, notes, habits, and memory. It is designed for people who want a contrarian AI assistant to pressure-test...

monolog

monolog is a note-taking tool that lets you write naturally without managing folders, titles, or tags. Its AI understands your entries, recognizes schedules and tasks, and reminds you when it matters. Semantic search finds past records by meaning....

Chiplab

Chiplab is an MCP server that lets AI coding agents build, run, and test embedded firmware on a virtual copy of the real chip, without physical hardware. It uses datasheets and run results to help agents write correct code and debug crashes, targe...

Click

Click is an MCP that provides research connectors to access context from professional platforms, social media, marketplaces, and financial sources. It installs in one minute within ChatGPT or Claude, expanding what users can do inside chat.

Prime Agent

Prime Agent is an open-source, self-improving coding harness for AI developers and researchers. It uses Recursive Language Model and Continual Harness abstractions to advance agent performance, achieving 95.5% on ARC-AGI-3 with Opus 5, above the r...

Open Index

Open Index provides a structured context layer for AI agents by defining entities and their relationships. It helps developers in domains like security and legal build agents that can navigate complex information graphs effectively.

Soup CLI

Soup CLI is a command-line tool for fine-tuning large language models on computers with limited GPU memory. It streams frozen base model layers from system RAM during LoRA training, allowing models that exceed GPU capacity to be trained. It is int...

Prompt Bridge

Prompt Bridge is a browser-based tool that captures AI chat threads from ChatGPT, Claude, Gemini, Grok, or DeepSeek, compresses them into a shorter context pack, and lets you paste that pack into another AI's input box. It runs locally in your bro...

Reference

Reference is a local embedding-based search tool that gives Claude quick access to past context and files. It indexes your documents into a compact, offline index, letting you or Claude retrieve relevant information instantly without loading a ful...

Gemini Robotics 2

Gemini Robotics 2 brings physical AI to robots, enabling them to interpret environments and reason through tasks. It is built for developers creating intelligent agents that adapt to physical spaces.

Meterless.ai

Meterless.ai lets you keep the work AI does during a chat session, not just the final answer. It is for anyone who uses AI for research, planning, coding, or file organization and wants to edit or rerun those tasks with cheaper models.

RecipeBook by Shofo

RecipeBook by Shofo lets AI teams build video training datasets from millions of public clips. Users search in plain language and filter by metadata to collect custom datasets for training video and multimodal models.

Freesolo Flash

Freesolo Flash trains small language models using SFT and RL through a coding agent. It quotes upfront pricing for the entire run. This allows engineers to build narrow task models within a set budget.

Lattics

Lattics is a desktop writing app for academic research and thesis writing. It combines a card-based knowledge base with AI tools for collaborative editing. Researchers use it to organize citations and analyze multiple documents locally.

Aymo AI

Aymo AI is an AI aggregator that gives teams access to multiple AI models in a single workspace. It replaces individual subscriptions with shared credits and lets users compare model responses side by side.

Kogvio

Kogvio is an AI tool that lets users highlight text on webpages and PDFs to ask questions and simplify concepts. It helps students and researchers read documents without leaving the page to copy and paste text into a chatbot.

BaseRT

BaseRT runs local language models for Apple Silicon users. This LLM runtime allows them to install and execute models directly on their own devices with a single command.

Ventorah

Ventorah performs computational fluid dynamics simulations for engineers using an OpenFOAM core and a built-in AI simulator. The software offers a simplified workflow to calculate fluid behavior and output accurate results.

Aye

Aye automates web research and content drafting in an AI browser for macOS and Windows. It helps professionals summarize pages and turn repeatable tasks into reusable skills while keeping sensitive steps under manual approval.

Breadcromb

Breadcromb is an AI browser that retains long-term memory of your web reading and documents. It helps researchers and writers by providing relevant context for tasks while keeping data private and allowing custom model selection.

Cito

Cito is a literature retrieval engine that searches academic papers for AI research agents. It provides a JSON API and native MCP endpoint so developers can build deep research tools without hitting rate limits.

VocalVia

VocalVia converts saved PDFs and articles into audio for readers who save documents but lack time to read them. Users can edit the script, assign voices, and adjust segments before generating the final audio.

Sourclip

Sourclip is a Chrome extension that captures and organizes research for NotebookLM users. It helps manage sources and converts NotebookLM Audio Overviews into private podcast feeds.

Clusy

Clusy is an AI-native notebook for data science that manages the whole pipeline. Its agent works alongside users from initial idea and environment setup to model training and deployment.

AlgoFly AI

AlgoFly AI manages computer vision datasets and annotation workflows for machine learning teams. This on-premise-first data platform keeps sensitive data on your own infrastructure instead of third-party clouds.

Are you in the Weights?

Are you in the Weights? queries many frontier and small models in parallel, clusters responses, and tells you how strongly they recognize you. It is for individuals investigating their traces in large language model training data.

Dump Memory

Dump Memory is a private second brain for professionals to save unorganized files, notes, and links. Users search by describing what they remember, allowing the AI to locate specific items and automatically generate related context.

ColibotAI

ColibotAI translates, summarizes, or explains selected text. It is for users who want to pick their own AI engine, including on-device or local models, ensuring nothing leaves their computer. Results save as searchable local notes.

OLO Robotics

OLO Robotics is a browser-based platform that lets you pick a simulator and robot, teleoperate, access ROS2 topics, get an AI coding assistant and use JavaScript/Python SDKs - deploy a Docker bridge to connect your robot in minutes.

OrchestraML

OrchestraML is a student-focused AutoML platform with agentic pipelines and human-in-the-loop controls. Its AI Audit Trail logs every decision in plain English so students can review and learn from the model-building process.

Verol

Verol checks AI responses by extracting claims from ChatGPT, Claude, and Gemini to verify them with live web lookups. It displays trust scores and source links locally in your browser. This tool is for researchers and coders needing verified facts.

Lium AI

Lium AI connects scientists, analysts, and data teams with terabyte-scale datasets to ask critical questions without hallucinations. It generates knowledge artifacts and turns ad hoc analysis into repeatable workflows.

TypingMind

TypingMind lets you use your LLM API keys to access models from 18 providers in one workspace-no extra subscriptions. Built for pro users with Projects, Fork/Parallel chat, Plugins and extensive customization.

Whisperstream

Whisperstream is a local dictation tool for Windows that transcribes speech on your CPU, so audio never leaves your device. It uses a push-to-talk hotkey to paste text into any window. It costs a one-time $29 payment with no subscription.

Hermes Desktop

Hermes Desktop is an open-source AI agent that grows with you. Native macOS, Windows and Linux app for automating workflows, managing tasks and research pipelines. By Nous Research.

MakersClaw

MakersClaw lets users hire AI employees for roles like support, sales, or research. Each runs continuously with its own memory and connects directly to Slack, Telegram, or Teams.

Web Clipper for NotebookLM

One-click web clipper for NotebookLM: save YouTube videos, articles, Reddit threads and Drive files as clean, queryable sources. Bulk import/delete, live Drive sync, duplicate notebooks and export to Anki/Obsidian/Markdown.

Clark

Clark provides developers and researchers an AI with its own cloud computer to complete async workspace tasks. Users submit a job and return later to inspect the resulting files, screenshots, and code patches.

Parastore

Parastore simulates synthetic shoppers in 3D store layouts: they browse, pick items, impulsively buy near checkout, and generate revenue and behavioral data. Open-source (MIT) framework for layout testing and agent-based retail research.

DecisionBox for Databricks

DecisionBox for Databricks connects to your workspace, auto-writes SQL, validates findings against your data, and delivers a ranked backlog of verified, prompt-free insights. Read-only, Unity Catalog scoped; supports Serverless, Pro, Classic SQL w...

Reader Alive

Reader Alive is an iPhone/iPad reader for EPUB, PDF, AZW3 and MOBI. Translate selected text or full chapters, listen with natural TTS, generate summaries and ask questions grounded in the book to tackle foreign-language, dense or technical reading.

Rixx

Rixx: search the web with verifiable answers backed by real citations. View images and charts inline, upload docs and images for multimodal queries, and organize research into Insights folders.

CatchAll by NewsCatcher

CatchAll by NewsCatcher: a web search API that converts open-web content into clean, deduplicated structured datasets. Submit natural-language queries to receive validated records ready for analytics, monitoring, or AI workflows.

Kept

Kept auto-saves AI chats to local .md files (Obsidian-ready), indexes them on your machine with an on-vault agent and 3D knowledge graph-search, link, and build on past prompts without accounts, cloud, or subscriptions.

Memdex

Memdex is a browser extension that saves AI chats to encrypted files on your laptop, detects relevant past discussions, and injects that context into any AI with one click-private, cross-model memory so you stop re-explaining.

Noteweave

Noteweave converts academic research into market-ready products using AI-driven research machines, helping businesses extract commercial innovations and build scientific moats from untapped papers.

Pipali

Pipali is a desktop AI coworker that automates work: research files and the web, draft briefs, spreadsheets, emails, reports and personal apps, run recurring tasks and monitor events with sandboxed permissions.

moar

moar optimizes documents for LLMs by stripping formatting bloat and redundant metadata, saving up to 95% of tokens. Local-only processing keeps files private and the optimization engine is free to use.

moar

moar optimizes documents for LLMs by stripping formatting bloat and redundant metadata, saving up to 95% of tokens. Local-only processing keeps files private and the optimization engine is free to use.

Draft

Draft is a privacy-first local workspace that captures AI chats via a browser extension, saves MHTML blobs to local browser storage, and converts them into editable, searchable notes for organizing and reusing AI answers.

Liminary

Liminary is a personal AI memory layer that stores files, web pages, videos, emails and chat transcripts so your AI tools use your curated context for consistent search, retrieval and writing.

Crin AI

Crin AI visualizes LLM internals with an animated node graph: see text → tokens → numbers → meaning as live data flows. Learn how models process language step by step, without diagrams or hand‑waving.

Gemini Deep Research Agent

Gemini Deep Research Agent automates long-horizon, multi-source research into fully cited, expert-quality reports. Combines open web and MCP proprietary data, analyzes PDFs/CSVs/media, creates charts and iteratively refines insights.

Subgrapher

Subgrapher organizes and shares knowledge as linked semantic subgraphs, letting references, research, messages and agents move across apps and the open web without defaulting to closed platforms.

SNEWPapers

SNEWPapers is the world's first AI-powered historical newspaper archive: 6M+ stories (1730s-1960s) with clean OCR, article summaries, semantic search and The Sleuth research assistant for fast, shareable historical discovery.

QuickCompare by Trismik

QuickCompare by Trismik tests your dataset across 50+ LLMs, giving side-by-side scores for quality, cost and latency, with no scripts required-so teams can pick the best model for their prompts.

ml-intern

ml-intern automates the ML research loop: finds papers, pulls datasets, runs GPU experiments on Hugging Face, iterates models, and produces research-backed models via CLI and web app.

ChatFolders

ChatFolders organizes AI conversations into folders across ChatGPT, Claude, Gemini and more. Local-first, lightweight browser extension with drag-and-drop, bookmarks and tags (5 free folders); upgrade for unlimited folders, search and sync.

Pioneer

Pioneer converts a plain-English task prompt into a fine-tuned model in under 10 minutes-auto-generating training data, selecting hyperparameters, training on cloud GPUs, evaluating vs frontier models, deploying and continuously retraining.

PangeAI

PangeAI lets AI agents automate geospatial work - coordinate reprojection, format conversion and spatial queries - so teams skip tedious GIS chores and get accurate maps and transformed datasets faster.

Katzilla

Katzilla is a single API for 250,195+ US government datasets, scrapes, watches and NL queries, returning structured citations (source URL, retrieval timestamp, SHA-256) so agents stop hallucinating and cite primary sources.

Arky

Arky is a free-form thinking canvas: place ideas, link and shape them into a hierarchy, then flip to a clean document. AI assists throughout; export to Markdown, Word or PDF.

AI Mode in Chrome

AI Mode in Chrome keeps search results and web pages side-by-side on desktop, so your research stays connected. Add open tabs, images or PDFs as context for AI answers grounded in those pages.

Innogath

Innogath converts scattered AI chats into a structured, non-linear research workspace: book-like reports, visual idea graphs, branching pages, linked notes and node-based chat to keep context and extend your work.

Reka Edge

Reka Edge is a 7B vision-language model for Physical AI. Its ConvNeXt V2 encoder uses 3× fewer image tokens for sub-second video analysis, low-latency object detection and agentic tool use on edge devices.

MiniAi

MiniAi - a menu-bar AI that explains selected text with one hotkey (⌥ Space). Receive concise, in-context definitions and explanations while reading papers or documents, so you stay focused without opening a browser. Free to try.

deckpipe.dev

deckpipe.dev converts any agent's MCP output into downloadable PPTX slide decks. Lightweight and agent-agnostic: let your AI create content while deckpipe handles file generation.

LayerProof Chromo

LayerProof Chromo reads your data, structures the story, builds real charts, and links every claim to its source-turning hours of copy‑pasting into minutes and delivering slides you can confidently present.

Bugsy

Bugsy helps you identify, collect and trade insects with inclusive features like VoiceOver and a phobia mode. Free collecting; premium AI chat and dynamic weather cards available.

Labelsets

Labelsets: a marketplace for labeled datasets with an automated Label Quality Score (7 checks). Objectively compare vendors, preview free 1,000-row samples by email, and pick higher-quality data for model training.

OpenRouter Model Fusion

OpenRouter Model Fusion runs prompts through multiple SOTA models, pre-evaluates their outputs, then a chosen fuse model synthesizes a single, controllable final answer. Mix open or closed models; free options let you test.

Google Gemma 4

Google Gemma 4 - an open (Apache 2.0) model family for advanced reasoning, multimodal inputs, 256K-context and agentic workflows. Offers frontier-level intelligence-per-parameter that runs efficiently on phones, laptops and local servers.

.

As seen on