Start here if you’re new. Otherwise:
- MODELS — The AI engines themselves. One at a time, directly. Models
- TOOLS — Apps built on top of an AI to do a specific job. You are here.
- ORCHESTRATION & AGENTS — Systems that choose or coordinate which AI answers you. Orchestration
- TERMS — The words you keep hearing, in plain language. Terms
Have a specific job to do? Go to Tools. Wondering why your AI seems to have gotten worse lately? That’s Orchestration.
Paste text into Napkin and you get an infographic. Paste the same request into Claude and you get a different infographic. Claude made both, but Napkin actually sent Claude a more instructions by people who thought hard about what makes an infographic readable.
Use the model directly when you know exactly what you want, but use a tool (like Napkin) when you do not have the words, expertise or time to ask for exactly what you want. You can often replicate the results directly with careful prompting, but these Tools are useful shortcuts. Often this means easier or task-specific interfaces and buttons for choices you forgot to make.
Most of the tools on this site are the same three or four AIs which means:
- Tools inherit the model’s errors along with its abilities: a grader built on a model that invents citations will invent citations.
- When the model changes, the tool changes, often with no announcement
- Your text travels: the tool’s privacy policy and the model company’s privacy policy are two different documents, and you need both.
Questions to ask:
- Which model is underneath? If they will not say, that is an answer.
- Does it train on what you or your students put in?
- What happens to your assignment when the underlying model changes? And it will. More on that in Under the Hood.
Understanding the variety of tools leveraging APIs on the market and which specific tasks each can perform for you, and which does a task better than another, is one element of the emerging AI literacy we all must develop. Note too that models are priced per million tokens, which is why long documents cost more. “Unlimited” usually means unlimited access to whichever model thetool company picked today. Bold entries are where I would start.
145 of 145 shown
Find & Understand
Literature Search
Worth knowing: Undermind, Consensus and Elicit all search the same underlying database — Semantic Scholar. They differ in workflow, not in what they can find. Undermind digs deepest and slowest, Consensus answers fastest, Elicit builds you a table. Elicit and Consensus add PubMed and OpenAlex on top.
- Undermind — Runs an iterative deep search that takes eight to ten minutes, using an LLM to judge relevance rather than matching keywords. Best when you want comprehensive coverage and can wait. Free with an edu account; scores how closely each paper matches your question.
- Consensus — Fast synthesis with a consensus meter showing where the literature agrees, and ranking that weighs study design and journal quality. Also available inside ChatGPT on any plan, including free — look in the plugin directory (see Terms on what ChatGPT means by "plugin").
- Elicit — Builds a table: one row per paper, columns you choose. Best when you need the same fields extracted across many studies. Adds OpenAlex and a clinical-trials corpus to the shared database.
- Storm — Stanford tool producing a Wikipedia-like report, writing from different disciplinary perspectives and showing its sources.
- Edison Scientific — Suite of scientific discovery agents: precedent search, literature synthesis, chemistry.
- Semantic Scholar — The free database underneath most of the tools below: 220M+ papers, built by the Allen Institute for AI. Search it directly, or let one of these tools search it for you.
- SciSpace — Broad suite including a paraphraser that rewrites or explains passages.
- NeuraLumi — Search tools with specialized algorithms for finding accurate scientific papers.
- Liner Scholar — Search across papers with cited answers; a lighter alternative when you want speed over depth.
Citation Analysis
- Scite — Extracts citations and analyzes whether they support or contradict the work.
Research Maps
- ResearchRabbit — Visual network of related articles. Like Spotify for research papers.
- Litmaps — Visual citation mapping of a literature.
- Connected Papers — Builds a visual graph of work related to one seed paper. Good for finding the shape of a field you do not know.
- Inciteful — Free citation network mapping. Paper Discovery builds a graph from one paper; Literature Connector shows how two papers relate.
Your Documents
- Gemini Notebook (renamed) — Formerly NotebookLM. Works only on documents you upload. Try it as a committee assistant: load every policy document and let faculty interrogate it.
- Mem — Chat with your own notes and data.
- Perplexity Spaces — Perplexity’s version of chatting with your own files.
- Claude Science — Integrates common research tools and packages, produces auditable artifacts, provides compute.
Data Analysis
- Julius — Computations and visualizations with your data; also writes reports and finds insights.
- Nomic Atlas — Best for analysis and visualization of huge unstructured datasets.
- Ailyze — Qualitative data and insights, with an AI interviewer for collecting data.
- Infography — Makes infographics from text.
Domain Specific
- OpenEvidence — Clinical AI copilot for point-of-care decisions. Usage figures are the company’s own.
Checked 2026-07-24
Teach & Assess
Grading
- GradeWiz — Extracts student names before grading, sends anonymous essays to multiple AIs, compares results.
- CoGrader — General grading and feedback with rubrics. Integrates with Google Classroom.
- Google Classroom AI feedback — Built-in AI feedback on written work.
- Snorkl — Reviews a range of work including images and provides feedback.
- Gradescope — Turnitin’s grading platform.
- Turnitin Feedback Studio — Feedback and originality checking.
- TimelyGrader — Imports assignments from your LMS and exports grades back.
- Graded.Pro — Free tier; integrates with Google Classroom.
- Brisk Teaching — Chrome extension that gives feedback inside Google Docs and shows a replay of how a document was written.
- Diffit — Takes any text or topic and produces a version at the reading level you specify, with questions and vocabulary.
Chatbot Assignments
- Parlay — Specialized chatbots for discussion and problems, with grading, in a protected environment.
- Mizou — Creates chatbots for problems or situations and grades student interactions.
- BoodleBox — Many models plus institutional controls: instructor visibility, Coach Mode, FERPA claims. The governance layer is the product.
- MagicSchool AI — Dozens of teacher-facing generators in one workspace: plans, rubrics, leveled texts, feedback. K-12 in origin but much of it transfers.
- Khanmigo — Khan Academy tutor built to guide rather than answer. Free for US teachers.
- Gemini Guided Learning — A Gemini mode built on the LearnLM family that walks students through problems instead of solving them. Free, and FERPA-covered inside Workspace for Education.
- SchoolAI — Teacher-controlled student chat spaces with monitoring of what students asked.
Study Assistants
- Nurovant — Turns lecture notes, PDFs or recordings into outlines, practice tests and flashcards.
- Turbolearn — Converts recordings and documents into study materials.
- GradeMaxx — Study materials from notes, PDFs and video.
- Scholarcy — Summarizes papers and documents.
- Mindgrasp — Summaries and transcriptions, marketed as learning 10x faster.
- CoralAI — Document summarizing aimed at reading faster.
- Caktus — Bills itself as a study assistant. Top tools include an essay writer, a humanizer, an AI-detector bypass and a discussion-board responder.
- Mapify — Concept maps from documents and topics.
- Albus — Visual exploration of new topics.
- Heuristica — Concept maps with templates for pro/con boxes, controversies and flash cards.
Capture
- Otter.ai — Transcribes, organizes and summarizes meetings.
- Fireflies — Meeting capture with analysis of who is talking.
- Fathom — Meeting notes and summaries.
- Zoom Companion — Zoom’s built-in note taking and summaries.
- Microsoft OneNote — Note taking with AI organization and summary.
- SpinachAI — Meeting notes combined with agent actions: creates tasks, sends updates, writes docs in 100 languages.
- Socrait — Voice class assistant that listens and summarizes what you said about the exam or assignment. Consent question worth asking.
Checked 2026-07-24
Make & Create
Infographics
- Napkin — Creates infographics and visualizations from your text with no prompting. Drop in the items and pick from the options.
- PaperBanana — Agentic framework for publication-quality academic illustrations.
- MyLens — Editable visuals from text, PDFs, video, webpages or spreadsheets. Data not used for training.
Images
- Imagen 4 — DeepMind’s image generator, available in ImageFX or Gemini.
- Krea — Built for aesthetic diversity rather than photorealism. Moodboards blend reference styles.
- Recraft — Interface for creating graphics with external models or its own. Good balance of power and practicality.
- Canva — Create images you can then edit. Upload brand standards and control style.
- Moda — Create and edit images with brand controls.
- Reve 2.0 — Fine-tuning tools for images; currently top-ranked.
- Ideogram — Top-ranked open text-to-image model. Strong with text in images.
- Flux.2 [klein] — Black Forest Labs. Fast and open source.
- Adobe Firefly — Trained on commercially safe licensed content, with access to Google, OpenAI and Black Forest Labs models.
- Seedream — ByteDance’s image model.
- Whisk — Use images as prompts for new images.
- Tess Design — Trained on licensed art; lets artists create and control their own style.
- Komiko — Creates comics, manga and animation.
- Midjourney — One of the best, used through Discord.
- Stable Diffusion — Good, cheap and open source.
Slides
- Gamma — Makes slides quickly from an outline or prompt.
- Beautiful.ai — Slide design with smart templates.
- Plus AI — Extension for making slides with AI inside Google Slides.
- PageOn — AI presentation builder.
- Claude Design — Design canvas that exports to Canva.
Video
- Veo — DeepMind’s video model, including dialogue and sound.
- Seedance 2.0 — ByteDance video model. Frighteningly good.
- Kling — Kuaishou’s model; cheaper and close to Veo.
- Runway — Gen4 plus Act-1, which lets you control animation with your own facial expressions.
- Hailuo — MiniMax’s video tool.
- Vidu — ShengShu Technology; ranked first in China.
- LTX — Lightricks. Open-source and very fast.
- Higgsfield — Motion control effects including overhead sweeps and camera moves.
- Hedra — Character video generation.
- Pika — Video generation including making a video with your younger self.
- Grok Imagine — Free and very good.
- Genie 2 — Converts images into interactive virtual worlds and games.
- Guidde — Turns a PDF, deck or screencast into a how-to video in multiple languages.
Voice
Before you sign up for any of these: ChatGPT now has a native voice mode that is nearly as good as the three in bold, and it is already on your phone. Try that first.
- Hume.ai — Understands your emotions and the emotion in what it says. HIPAA compliant.
- Sesame — Very natural voice-to-voice conversation. Good for practicing conversations.
- ElevenLabs — Text to audio, audiobooks, voice clones and licensed voices. Powers the podcast feature in Gemini Notebook.
- Cartesia — Suite of voice tools including cloning.
- Murf — Text to speech and conversational voice agents.
- Kokoro — Open-source natural speech from text.
- Octave — Hume’s text-to-speech that understands what it is saying and adds emotion.
- Scribe — Speech to text in 99 languages.
- Amazon Nova Sonic — Understands both what you say and how you say it.
- Chatterbox — Resemble AI’s open-source model. Clones a voice from five seconds of audio; can run entirely offline.
- ChatGPT Voice — Native voice conversation in the app you probably already have. Not as expressive as Hume or Sesame, but close enough that it is the right place to start.
Music
- Suno — Turns instructions into a new song. Try a gratitude song, or walk-on music for your class.
- Udio — Song generation from text.
- Producer — Formerly Riffusion. More compositional complexity.
- Mureka — Song generator that can start from an uploaded melody.
- ElevenLabs Music — Trained on licensed data and cleared for commercial use.
- Lyria — Three-minute songs with structure, plus a live mode for jamming.
Music Education
- Music Teacher (Jenova) — Theory, technique, ear training, sight-reading, improvisation and practice strategies.
- ABC Transcription Tool — ABC notation is shorthand music notation for computers, handy when asking AI for music notation.
- Moises.ai — Stem separation, voice studio, video recording and an AI studio.
- Fadr — Remixes and extracts vocals.
- Mooz — Video conferencing for music education with MIDI and zero-latency collaboration. Includes MIA for exercises.
Avatars
- HeyGen — Lip-synced avatar from a photo and a script. Can attach a knowledge base to an interactive avatar.
- Wan — Video avatars.
- Synthesia — Interactive video and voice clones.
- Colossyan — Video avatars and voice clones.
Writing
- Grammarly — Known to students before AI. Blurs the line between improving and cheating.
- Quillbot — Paraphrasing and rewriting.
- Lex — Try this first for your own writing.
- NotionAI — Writing inside Notion.
No Code
- Base44 — Reliably turns ideas into working apps.
- Replit — Browser-based editing, AI generation, instant deployment, many languages.
- Lovable — Easy interface aimed at people who are not developers.
- Cursor — Desktop tool that understands full codebases and does multi-file edits.
- GitHub Copilot — Code completion integrated into popular editors.
Checked 2026-07-24
Practice & Converse
Language
- Duolingo — First out of the gate with Duolingo Max.
- Speakology — Video conversation partners and oral exams in dozens of languages, built for schools.
- Langua — More engaging than Duolingo for conversation practice.
- Glossarie — Immersive language practice.
- TalkPal — Conversational language practice.
- Univerbal — Combines text and audio; audio-only mode is better for real conversation.
- Speakable — All-in-one language TA.
- DeepL — Translation, often better than Google Translate.
- LingoSub — Adds subtitles in any language to videos.
- DeepDub — Dubs speech into another language keeping the original voice, with lip-sync.
- Rask — Dubbing with voice preservation and lip-sync.
Accessibility
- Be My Eyes — Describes images, navigates interfaces and reads web content aloud.
- Seeing AI — Microsoft app that describes the visual world aloud.
Companions
- Character.ai — Roleplay companions. Documented wellbeing risks for younger users.
- Replika — AI companion. Documented wellbeing risks for younger users.
- ElliQ — Companion focused on senior loneliness.
- EdSights — Talks to students as a way for colleges to hear student voices.
Checked 2026-07-24
Nothing matches that. Try fewer filters.