FullStack Alpha logo BestAITradingTools.com by FullStack Alpha
Red and green candlestick chart recovering after a pullback on a dark background

Open-source AI trading directory

Best AI Trading Repos on GitHub Right Now | 35 Scored Before You Trust One

GitHub is where the free AI trading tools live. It is also where abandoned code, copycat bots and key-stealing traps live.

We checked 35 of the best-known AI trading repos, verified every star count and update date, and scored each one on what real traders actually do with it. No code required to read this page.

Listen · 1:28 audio briefing · voice: EricNew to GitHub Trading Bots? Start Here | 90 Seconds, No Code
Read the transcript

Somebody told you the best AI trading tools are free on GitHub. They're right. They just left out the part that matters. Here it is in ninety seconds.

GitHub is where programmers share code. Microsoft owns it, and more than a hundred million developers use it. Each project is called a repo. Think of it as a recipe card anyone can copy.

The first number you'll see is stars. A star is a bookmark, not a review. TradingAgents has around a hundred eighteen thousand stars, and the people who read its code still found real problems.

So we built the Community Score. Five checks out of a hundred: how many people use it, whether anyone still updates it, what traders say on Reddit and Hacker News, how many builders work on it, and whether the license lets you use it. Freqtrade tops our list at ninety-six.

Here's the move that saves beginners money. Never start with real cash. Most serious tools have a fake-money mode. Freqtrade calls it dry-run. Hummingbot has paper trading. And Alpaca gives you a free paper account, so you can watch a bot trade for weeks before it touches a dollar.

Your system: pick one repo, run it on fake money, and keep notes.

Scroll down for the top ten, the beginner glossary and the full directory of thirty-five.

This is education, not financial advice. Your market edge starts with the right tool. Stay alpha.

repos verified
35
combined stars
689k
deep dives
27
data checked
2026

FullStack Alpha Editorial Team

Reviewed by Jay Rocco, Chief Editor

New here? Start with this

What Is a GitHub Trading Repo, in Plain English?

Think of GitHub as a giant public library of free software. A trading repo is one book on the shelf: the code for a bot, a backtester, or a plug that lets AI read the market. Some are brilliant. Some have not been opened in years.

  1. Step 01

    Pick a lane

    Want an AI to talk to your broker? A bot that trades crypto? A way to test ideas on history? Each lane below answers one of those.

  2. Step 02

    Check the score

    Every repo carries a Community Score out of 100. It tells you if the project is alive, trusted and legal to use, before you install anything.

  3. Step 03

    Start in paper mode

    Run it with fake money for weeks. A strategy that cannot survive paper trading will not survive your real account.

The 8 words you need to read this page

Repo
Short for repository. A project's folder of code on GitHub, free to look at.
Star
A bookmark. Lots of stars means lots of interest, not proof it works.
Fork
Someone's own copy of the code that they build on. A sign people use it.
Backtest
Replaying a strategy on old prices to see what would have happened.
Paper mode
Trading with fake money against real prices. Always start here.
API key
A password your code uses to talk to a broker. Guard it like your PIN.
MCP server
A plug that lets an AI assistant like Claude read data or use an app for you.
License
The rules for using the code. No license means no clear right to use it.

See It in One Picture | Reading a Repo and the Safe Path to a First Trade

Two cheat sheets for visual learners. Save them, share them, come back to them.
Infographic: how to read a GitHub repo. Stars are bookmarks, not reviews. Forks are copies people build on. Last commit shows if anyone is still maintaining it. MIT or Apache licenses are the safest. Issues are where users report bugs. The README is the instruction manual.
Every number on a GitHub page, translated. Example repo shown, not a real project.Full size ↗
Roadmap infographic in 7 steps: pick your lane, check the Community Score, run the 10 safety checks, backtest on history, paper trade 2 to 4 weeks with an Alpaca paper account or Freqtrade dry-run, go live small with trade-only API keys, review every week.
Seven steps from a repo you just found to a system you trust. Step 5 is the one people skip.Full size ↗

Which Kind of AI Trading Repo Do You Need?

Six lanes. Pick the job you want done, and skip the rest.

Which AI Trading Repos Score Highest Right Now?

The top 10 by Community Score. Stars tell you who showed up. The score tells you who stayed. Want the most popular instead? See the 10 most-starred repos explained.
#RepoWhat it doesStarsScore
1Freqtradefreqtrade/freqtradeTrading Bots56,11096Score
2Nautilus Tradernautechsystems/nautilus_traderBacktesting31,01792Score
3Qlibmicrosoft/qlibQuant and ML49,79190Score
4CCXTccxt/ccxtResearch and Data44,61490Score
5Hummingbothummingbot/hummingbotTrading Bots20,77587Score
6QuantConnect LEANQuantConnect/LeanBacktesting22,28887Score
7Machine Learning for Tradingstefan-jansen/machine-learning-for-tradingQuant and ML21,62487Score
8TradingAgentsTauricResearch/TradingAgentsAI Agents117,97585Score
9OpenBBOpenBB-finance/OpenBBResearch and Data74,84585Score
10edgartoolsdgunning/edgartoolsResearch and Data2,82685Score

Prefer to Listen? Audio Briefings on the Top Repos

90-second plain-English briefings on 25 of the top repos: what each repo does, the five things we like, what to watch out for, and who it is for.

Browse All 35 AI Trading Repos

Filter by the AI you use, your skill level and what you trade. Every card shows the repo's real GitHub page, its score, and what it plugs into.
Works with
Skill level
Market
Type

Showing 35 of 35 repos

Which AI Assistants and Brokers Do These Repos Connect To?

AI assistants

Claude, ChatGPT and Codex plug in directly through MCP servers. Agent frameworks let you swap models, including free local ones. Grok has no built-in connector on these 35 yet, but its API speaks the same format many frameworks accept.

ClaudeChatGPTCodexGeminiDeepSeekLocal modelsGrok

Brokers, exchanges and data

Where the code actually sends orders or pulls numbers. Check each deep dive for the exact list, since support varies repo to repo.

AlpacaInteractive BrokerstastytradeTradingViewBinanceCoinbaseKrakenOKXBybitSEC EDGARMetaTrader 5

How Do You Spot a Poisoned GitHub Trading Repo?

Fake trading bots are a known way to steal API keys and crypto wallets. They copy a real project's name, buy stars, and wait.

A 5,000-star repo created last week is not a success story. It is a warning sign.

Run the 10-point safety check

Our Community Score blends five public signals: stars, recent updates, builder activity, license and what traders really say. No vendor pays for placement.

It measures whether a project is alive and trusted. Nothing on GitHub can promise profits, and nobody honest will.

See how the score works

Keep reading across FullStack Alpha

Not ready to run code yourself? Our hands-on AI stock tool reviews on AIStockPickerApp.com score finished products on five pillars, the AI stock tools directory at AIStockPickerApps.com sorts 200+ tools by the job they do, and AIStockTradingBots.com walks through broker APIs, bot setup and backtesting step by step. Or head back to the Trading Workflows home page to match tools to your own routine and budget.

Best AI Trading Repos FAQ

What are the best AI trading repos on GitHub right now?

By our Community Score, Freqtrade, Nautilus Trader, Qlib, CCXT and QuantConnect LEAN lead the pack. The best one for you depends on your lane: Freqtrade for crypto bots, LEAN or Nautilus for serious backtesting, and an MCP server like Alpaca's if you want Claude or ChatGPT to talk to your broker.

Are AI trading bots on GitHub safe to run?

The well-known projects are, if you start in paper mode and keep API keys out of the code. The danger is copycat repos built to steal keys. Check stars against the repo's age, read before you run, and never paste a live broker key into code you have not read.

Do I need to know how to code to use these repos?

For most of them, a little Python helps. The MCP connectors are the exception: you talk to Claude or ChatGPT in plain English and the repo handles the code. Look for the Beginner filter in the directory.

Which GitHub trading repos work with Claude, ChatGPT or Codex?

Use the Works with filter. MCP servers like Alpaca's connect straight to Claude, ChatGPT and Codex. Agent frameworks like TradingAgents let you pick the AI model, including free local models that run on your own computer.

Can I use Grok with these trading repos?

None of the 35 lists Grok as a built-in option on its main page today. Several agent frameworks accept any model that speaks the OpenAI format, and xAI built its API that way, so technical users can wire Grok in. Expect to edit a config file.

Is open-source trading software really free?

The code is free. Running it is not. Budget for market data, a broker account and somewhere to host it. Every deep dive lists what you need to run that repo.

What is the FullStack Alpha Community Score?

A rating out of 100 built from public data: GitHub stars, how recently the code was updated, how many people built on it, license clarity and what traders say on Reddit and Hacker News. It is not an Alpha Score, which only comes from our own hands-on testing.

Incoming transmission | Free weekly newsletter

Stop paying for tools that don't earn their seat.

FullStack Alpha has tested 200 AI stock tools. Be honest: how many of yours do the same job, and which one did you forget you signed up for?

One email a week fixes that: which tools are worth the money, what they cost right now, and how the traders on this site wire them into a routine. Five minutes to read. Zero noise.

Free. One email a week. Unsubscribe in one click. We never sell your email.

  • Hot stock picks
  • Paid signal pitches
  • Fake countdown timers
  • Gurus selling courses

Inside every issue

5 MIN READ

  1. 01

    The Verdict

    One AI trading tool tested and scored. What it does well, what it costs, and the catch the sales page leaves out.

  2. 02

    Price Watch

    Price changes and live promos on the tools traders actually use, monthly and annual side by side, so you stop overpaying.

  3. 03

    Workflow Teardown

    One educator's routine taken apart: the scan, the chart, the tools, and the free swap for every paid one.

  4. 04

    Scan of the Week

    Exact scanner settings you can copy, pulled from the educator's own lesson with the timestamp.

Process beats prediction. This is the email that keeps your process honest.

FullStack Alpha logo

FullStack Alpha Network