TerraByte-Dev

TerraByte-Dev / hinge-auto

Public

Emulator + ADB + Claude/Ollama vision: experimental Hinge automation that judges profiles and writes personalized openers against a user-defined rubric.

14
6
50% credibility
Found May 19, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

HingeAuto is a tool that automates swiping on the Hinge dating app using AI. It runs on a virtual Android phone on your computer, takes screenshots of each profile, asks an AI to judge whether you'd be interested based on your preferences, and either skips or sends a like with a personalized opener message. The project includes a profile-scanning feature that reviews your own profile and suggests improvements. The README explicitly warns that this violates Hinge's Terms of Service and could result in account bans, recommending use only on a throwaway account.

How It Works

1
👋 You hear about a dating app automation tool

A friend mentions HingeAuto — a project that uses AI to swipe through Hinge profiles for you, judging each one against your preferences and sending personalized openers.

2
📱 You set up an Android phone on your computer

You download Android Studio and create a virtual phone that runs on your computer. This acts as a fake phone where the app can run automatically.

3
🔌 You connect an AI assistant to think for you

You connect either a paid AI service or a free one running on your own computer. This AI looks at each profile screenshot and decides if you'd be interested.

4
⚙️ You tell it what kind of person you're looking for

You write a short description of your ideal match — your preferences, age range, deal-breakers — and pick a message style that matches your personality.

5
🚀 You start the automation and watch it work

With Hinge open on your virtual phone, you launch the bot. It scrolls through profiles, shows each one to the AI, and either skips or sends a like with a personalized message.

6
You can also analyze your own profile
📸
Review your photos

Get feedback on which photos work best and which ones look like duplicates

✍️
Improve your prompts

See which answers are too generic and get rewrite suggestions

🎉 Your dating app runs on autopilot

The bot works through your daily likes while you do other things, sending messages that sound like you to people the AI thinks you'd match with.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is hinge-auto?

Hinge-auto automates swiping on Hinge by driving an Android emulator through ADB. For each profile, it scrolls through the photos and prompts, screenshots the frames, sends them to a vision model (Claude or Ollama) with your personal rubric, and either skips or taps the heart and types a tailored opener. You define preferences in a mode file, pick a voice template for message tone, and optionally set premade openers the model can use verbatim. The project runs as `python main.py` against a connected emulator, with a dry-run mode for testing rubrics without spending your daily like cap. There's also a self-profile scanner that reviews your own profile for improvements without touching Hinge's terms of service.

Why is it gaining traction?

The hook is the vision-plus-LLM loop on a phone UI. Most dating automation tools are click-timers or simple keyword filters; this one actually reads profiles the way a human would and writes openers referencing specific photos or prompts. The mode system lets you encode personal preferences without touching code, and the two backend options (Anthropic for quality, Ollama for free) serve different risk tolerances. The self-scan feature is a clever sidestep for the one thing that doesn't violate ToS.

Who should use this?

Developers curious about AI agents operating on real consumer UIs. Dating-app power users with Hinge+ who want to optimize their like efficiency. Anyone willing to treat a throwaway account as an educational toy. Not suitable for production use on a real account given ToS risk and the project's experimental maturity.

Verdict

At 14 stars with minimal test coverage, this is firmly in the "interesting experiment" category. The documentation is thorough and the architecture is thoughtful, but the 0.5% credibility score reflects genuine immaturity. Try it on a burner emulator with dry-run enabled, but don't bet a real Hinge account on it yet.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.