Swipe right on decisions.

📊 Data files 🎬 Movies & TV 📈 Prediction markets 🧠 Flashcards 🔌 Any API

Turn any list into a swipeable deck. Sort 100 items in 2 minutes. Track prediction markets. Study with flashcards. Connect live APIs. Export to Excel. Works on any device, no install needed.

↑

Drop file here or click to upload

Supports Excel, CSV, JSON, TSV, Markdown, TXT

XLSX CSV JSON MD TXT

What to Watch, Listen & Read

Live charts from iTunes & Open Library. Updated daily. Swipe to build your watchlist.

Prediction Markets & Web Search

Real-time prediction markets from Polymarket — see where whales put money. Pair with web search to research any topic.

Study Mode

Spaced-repetition flashcards — swipe to learn. Built-in decks for markets, stocks, and AI.

Quick Start

Developer Examples

Live Datasets

Public API data — products, recipes, movies, startups. Real data, instant.

Learn more

About SwipeBase +

Tinder for everything except dating. Turn any list into a swipeable card deck. CVs, expenses, tasks, products, events, research - whatever you have. Right = Yes, Left = Pass, Down = Maybe.

Every competitor (Reelgood, Chicken Tender, Sorce, Sephora, Mailbox) locks into one vertical. SwipeBase works with any data type.

No account No install Privacy-first Mobile + Desktop AI-ready
How it works +
1

Paste anything - files, text, URLs, or AI output. Auto-detects emails, CVs, tables, JSON, CSV, Markdown.

2

Swipe cards - Right = Yes, Left = Pass, Down = Maybe. Tap to expand. 50 items in 3 minutes.

3

Export - Yes/Maybe/Pass piles to Excel, CSV, or JSON.

Why swiping works - the science +
Hick's Law - Decision time grows with options. One card = one fast choice.
Paradox of Choice - More options = less satisfaction. Swiping forces satisficing.
System 1 (Kahneman) - Swiping = fast intuitive processing. Gut first, analyze later.
Zeigarnik Effect - Unfinished tasks create tension. Card counter pulls you to complete.
Cognitive Load - Binary input = minimal mental effort. No spreadsheet needed.
Dopamine - "What's next?" = same loop that makes Tinder addictive. For productivity.
Use cases +
HR 7.5M swipes
Products 2x impact
Email 1M waitlist
Shopping 3-5x conv.
Expenses swipe to approve
Tasks daily triage

Your data never leaves your browser. No server, no tracking, no accounts.

APIPOOL + For Builders & AI Agents

Smart API endpoints powered by APIPOOL — describe what you need, get swipeable cards back

For AI Agents +

SwipeBase is agent-native. Any AI agent can create decision decks, load data, and read results via JavaScript API, postMessage, or standard discovery protocols.

JavaScript API

SwipeBaseAPI.loadData(items, "Deck Name")
SwipeBaseAPI.getResults() // {yes:[], maybe:[], pass:[]}
SwipeBaseAPI.getState() // {total, decided, remaining}

iframe Embedding

postMessage({type:'swipebase:load', items, name})
// listen for 'swipebase:results'

Supports A2A protocol (v0.3.0), OpenClaw skills, MCP integration, and ERC-8004 registration. Full docs: AGENTS.md

Every card includes an api_source field — visible on export so you can trace which endpoint produced each result.

Deck Name

0 items

0 Yes
0 Maybe
0 No
PASS
YES
← Pass · → Yes · ↓ Maybe · Space Skip · Tap card to expand

Live Results

0 left

No items yet

User

email@example.com

0
Decks
0
Swipes
0%
Yes Rate

Deck History

Review Complete

Deck Name

0
Selected
0%
0
Maybe Later
0%
0
Passed
0%
0
Total Reviewed
-

Distribution

✓ Selected 0

? Maybe 0

Export All

Saved
APPROVED
Decision title

Import AI-Generated Decisions

Paste a JSON array of decisions generated by AI. Use the AI Card Factory "Decision Hub" tab for a ready-made prompt. Supports markdown code blocks.

Decision Hub

Start typing to see live preview

AI Prompt Guide for SwipeBase

Use these prompts with any LLM (ChatGPT, Claude, Gemini) to convert your data into perfect SwipeBase JSON. Copy a prompt, paste your data, and get a JSON file ready to import.

SwipeBase JSON Rules

1. Output must be a JSON array: [{ ... }, { ... }]
2. Every item MUST have a "name" or "title" field (card title)
3. Use "description" for the main text body
4. Use "category" or "type" for color-coded badges
5. Use "url" for clickable links
6. Use "image" for card images (full URL)
7. Use "priority" for importance (HIGH, MEDIUM, LOW, CRITICAL)
8. Use "amount", "price", "salary" for highlighted numbers
9. Use "deadline" or "date" for time info
10. Use "location" for place info
11. NEVER lose data - every field from source becomes a JSON key
12. Long text (>100 chars) gets its own section
13. Short text (<100 chars) becomes a labeled field
14. Arrays should be joined with " | " separator
15. Nested objects should be flattened to "parent_child" keys

Export Results