Freqtrade
Free
freqtrade
Core parameters and statistics
| Project | Specifications |
|---|---|
| Project Name | Freqtrade |
| Category | Quantitative Trading / Cryptocurrency |
| Open Source License | GPL-3.0 |
| GitHub Stars | 52.3k |
| Programming Language | Python 3.11+ |
| Delivery form | Self-hosted / CLI / Docker / Web UI |
| Target users | Developers, quantitative traders, asset management teams |
| Core sponsors (if any) | Community driven (no single sponsor) |
The number of GitHub stars, forks, and contributors to an open source project are key indicators of community activity. The license type determines the freedom for commercial use and redistribution.
User and market recognition
Freqtrade has a broad user base in the field of cryptocurrency quantitative trading. GitHub 52.3k Star, 10.9k Fork, Discord community is extremely active. It is included in the Journal of Open Source Software (JOSS) and has academic citation value. Publicly available data shows that 130+ projects and enterprises use Freqtrade in production environments. Obtain official or community recognition from 15+ mainstream exchanges such as Binance, Bybit, OKX, Kraken, etc.
Cost advantage
| Cost Dimension | Description | ||
|---|---|---|---|
| Software license | Free (GPL-3.0 open source) | $25~$100/month (commercial platform) | High development manpower |
| Server cost | About $10~$30/month (cloud instance) | Included in subscription fee | Depends on architecture complexity |
| Strategy Development | Community Free Strategy Template | Restrictions | Need to be developed from scratch |
| Backtest optimization | Built-in free engine | Partially paid | Need to build by yourself |
The core cost advantage of the open source model is zero license fees + complete data sovereignty, but you need to bear the deployment and operation and maintenance costs yourself. Comparing the TCO with SaaS solutions requires comprehensive consideration of the three dimensions of software fee + infrastructure + manpower.
Architecture and core capabilities
- Architecture Overview: Event-driven asynchronous architecture, an event loop engine built based on Python asyncio, ensuring low-latency execution under high-frequency market conditions. CCXT library unifies abstract exchange API differences.
- Core Engine: The backtesting engine supports custom time periods, transaction fee models and slippage simulations; the real offer engine supports limit orders, market orders, and stop-loss orders, and is compatible with spot and contract transactions.
- Extension mechanism: Strategy code and execution logic are completely decoupled. Users only need to implement callback functions such as
populate_indicators()to define transaction logic. The FreqAI module provides adaptive predictive modeling. - Deployment Options: Officially provides one-click deployment of Docker images, supporting
develop(development) andstable(stable) dual branch strategies.
Model and version evolution
| Version | Date | Key Changes |
|---|---|---|
| 2026.6 | 2026-06 | FreqAI model saving improvements, Bybit EU support, Python 3.11+ adaptation |
| 2026.5 | 2026-05 | New engine for hyperparameter optimization and enhanced backtest performance report |
| 2026.4 | 2026-04 | Web UI reconstruction, real-time position dashboard |
| 2025.12 | 2025-12 | Hyperliquid DEX support, multi-timeframe strategy analysis |
Version records are subject to the GitHub Releases page.
Technical advantages
- Architecture Design: Event-driven asynchronous architecture + CCXT unified exchange interface, one-time development and deployment of multiple exchanges.
- Performance: FreqAI adaptive predictive modeling, built-in feature engineering pipeline and label generator. The hyperparameter optimization engine supports multi-objective optimization.
- Safety Design: Multi-layered risk control architecture - from global (daily transaction limit, maximum drawdown) to position level (stop loss, take profit, trailing stop).
- Observability: FreqUI Web console provides real-time position monitoring, yield curve, and transaction flow. Support Telegram Bot remote control.
Deployment pitfall guide
Based on community deployment experience, the following frequently asked questions are:
1. Exchange API permission configuration: The API permission configuration of different exchanges varies greatly, and some exchanges need to enable contract trading permissions separately. Solution: Refer to the configuration guide of each exchange in the official documentation. The API Key only grants trading and reading permissions, but does not grant withdrawal permissions.
2. Data quality issues: There is a big difference between the backtest results and the actual performance, usually due to insufficient historical data accuracy or inaccurate transaction fee models. Solution: Use officially supported exchange data sources and enable slippage simulation and transaction fee models during backtesting.
3. Strategy overfitting: Hyperparameter optimization overfits historical data, resulting in poor real performance. Solution: Use cross-validation and out-of-sample testing, and the backtesting cycle covers different market environments (bull market/bear market/concussive market).
How to use
| Entrance | Installation/Usage |
|---|---|
| Docker (recommended) | docker pull freqtrade → Create user data directory → Configuration → Start |
| Source code installation | pip install freqtrade → Create configuration → Download data → Run |
| Web UI | FreqUI console, no command line required |
Typical steps to get started quickly: freqtrade create-userdir → freqtrade new-config → freqtrade download-data → freqtrade backtesting → freqtrade trade (dry run mode first).
Product pricing (open source projects are usually free)
| Tiers | Price | What's Included |
|---|---|---|
| Open source core | $0 | All core functions (backtesting, live trading, FreqAI, Web UI) |
| Cloud hosting version | None | Users need to deploy their own servers |
| Enterprise Edition | None | Community Self-Service Support |
Users are responsible for their own server fees (approximately $10~$30/month) and exchange fees.
Application scenarios
- Personal Quantitative Trading: Individual investors deploy automated strategies to achieve 24/7 uninterrupted trading, and reduce real trial and error costs through backtesting and advance screening strategies.
- DeFi Market Making and Arbitrage: Supports DEX/CEX exchanges such as Hyperliquid and Gate, and can be used for liquidity mining and cross-exchange spread arbitrage.
- Quantitative strategy research and development: The research team uses backtesting engines and hyperparameter optimization tools to quickly iterate strategy assumptions, covering the entire link from research to actual trading.
Applicable people
- Developers: Python developers can deeply customize strategies and access FreqAI for machine learning strategy development.
- Operation and Maintenance/Platform Team: Small quantitative teams can use Freqtrade as a strategy execution gateway to achieve unified management of multiple strategies and multiple exchanges.
- Technical Decision Maker: Evaluate technology selection - Freetrade is suitable for individuals and teams with technical capabilities, but is not suitable for users with zero foundation in programming.
- Unsuitable Boundary: Enterprise users who need official SLA guarantee (the community provides self-service support); high-frequency trading scenarios are limited by Python performance.
Comparison of competing products
| Comparison Dimensions | Freqtrade | 3Commas | HaasOnline |
|---|---|---|---|
| Open Source/Closed Source | Open Source | Closed Source | Closed Source |
| License Type | GPL-3.0 | Commercial | Commercial |
| Community activity | 52.3k Stars | — | — |
| Deployment complexity | Medium (requires self-deployment) | Low (SaaS) | Low (SaaS) |
| Enterprise features | No official commercial support | Enterprise version available | Enterprise version available |
Summary and Outlook
Freqtrade has become one of the most popular solutions in the field of cryptocurrency quantitative trading due to its advantages of being open source and free, with an active community and comprehensive functions. The ecosystem of 52,000 GitHub Stars, 340+ contributors and 130+ enterprise users proves its maturity.
Core Advantages: Completely free, FreqAI machine learning integration, 15+ exchange support, backtesting + optimization + full link coverage of real trading. Current limitations: There is no official commercial support channel, advanced strategies require Python programming capabilities, and high-frequency trading scenarios are limited by Python performance. Development risks lie in community sustainability and the adaptation costs caused by exchange API changes. It is recommended to evaluate community activity trends and core maintainer stability before adoption.
| Parameter item | Detailed description |
|---|---|
| Development Language | Python 3.11+ |
| Code repository | GitHub (github.com/freqtrade/freqtrade) |
| Open Source License | GPL-3.0 |
| GitHub Star | 52.3k |
| GitHub Fork | 10.9k |
| Total Contributors | 340+ |
| Enterprise users | 130+ (public projects) |
| Release version | 115+ Release |
| Supported exchange types | Spot + Contract |
| Data persistence | SQLite |
| Minimum hardware requirements | 2GB RAM, 1GB disk 2vCPU |
| Operating System | Windows / macOS / Linux |
| Docker support | Official Docker image provided |
Freqtrade has received a total of 52,000 GitHub Stars and 10,000 Forks, making it one of the most active open source trading robot projects on GitHub. The project has more than 340 contributors, has released more than 115 official versions, and community discussions are active on the Discord server.
User and market recognition
Freqtrade has a wide user base and community recognition in the field of cryptocurrency quantitative trading.
Community size: GitHub 52.3k Star, 10.9k Fork, Discord community is extremely active, with hundreds of technical discussion posts every day.
Academic citation: Freqtrade is included in the Journal of Open Source Software (JOSS), and the DOI can be cited, proving its academic and technical value.
Enterprise adoption: Publicly available data shows that 130+ projects and enterprises use Freqtrade in production, covering quantitative funds, individual traders and DeFi research teams.
Exchange Ecosystem: Obtained official or community recognition from 15+ mainstream exchanges such as Binance, Bybit, OKX, Kraken, etc., and supports access to long-tail exchanges (through CCXT library expansion).
Cost advantage
Freqtrade itself is completely free and open source, and users only need to bear the server running costs and exchange transaction fees.
| Fee items | Freqtrade | Commercial quantitative platform (such as 3Commas) | Self-developed trading system |
|---|---|---|---|
| Software license fee | Free (GPL-3.0) | $25~$100/month | High development labor costs |
| Server cost | About $10~$30/month (cloud instance) | Included in subscription fee | Depends on architecture complexity |
| Strategy development cost | The community provides free strategy templates | There are restrictions, more fees are required | Requires development from scratch |
| Backtesting and optimization | Built-in free engine | Some features require paid upgrades | Need to build a self-built backtesting framework |
| Machine learning integration | FreqAI built-in for free | No or additional purchase required | Professional ML engineer required |
For individual traders, using Freqtrade only requires a low-configuration cloud server (about $15/month) to run 24/7 automated strategies, and the total cost is much lower than commercial SaaS or self-developed solutions.
Main functions
- Backtesting Engine: Strategy backtesting based on historical OHLCV data, supporting custom time periods, transaction fee models and slippage simulations. Users can quickly verify the effectiveness of the strategy and avoid firm offer risks.
- FreqAI Machine Learning Integration: Built-in adaptive predictive modeling module, which can automatically train the model and adapt to market changes in real time. Supports TensorBoard visualization and model persistence, suitable for quantitative researchers to quickly build ML strategies.
- Hyperopt): Provides a genetic algorithm to search for optimal strategy parameters, supports multi-objective optimization (Sharpe ratio of return, maximum retracement, etc.), and improves strategy tuning efficiency several times.
- Real-time transaction execution: Based on an event-driven asynchronous engine, it supports multiple order types such as limit orders, market orders, and stop-loss orders, and is compatible with spot and contract (including leverage) transactions.
- Multi-exchange support: Access 15+ exchanges through the CCXT library, support mainstream platforms such as Binance, Bybit, OKX, Kraken, Gate, Hyperliquid, etc., and can be extended to long-tail exchanges.
- Web UI Management Panel: Built-in FreqUI web console, providing real-time position monitoring, yield curve, transaction flow and strategy management interface, no technical background is required to operate the system.
- Telegram Remote Control: Supports Telegram Bot remote command management, including starting/stopping transactions, viewing positions, manually closing positions, viewing income reports, etc., to facilitate mobile monitoring.
- Dynamic currency selection and risk control: Supports dynamic whitelist/blacklist screening based on trading volume, volatility and other indicators, and has built-in risk control mechanisms such as maximum drawdown protection, daily transaction limits, stop loss and take profit.
- Data Download and Management: Provides command line tools to download, convert and manage historical K-line data, supporting multiple data formats (JSON, Pandas, Feather, etc.).
Model and version evolution
Freqtrade's version number uses the year.month format (such as 2026.6), and 10~12 incremental versions are released every year.
| Version | Release Date | Key Changes |
|---|---|---|
| 2026.6 | 2026-06 | FreqAI model saving improvement Bybit EU supports Python 3.11+ adaptation |
| 2026.5 | 2026-05 | New engine for hyperparameter optimization and enhanced backtest performance report |
| 2026.4 | 2026-04 | Web UI reconstruction, real-time position dashboard |
| 2025.12 | 2025-12 | Hyperliquid DEX support, multi-timeframe strategy analysis |
| 2025.6 | 2025-06 | FreqAI RL (reinforcement learning) strategy framework launched |
| 2024.12 | 2024-12 | Full support for contract/margin trading |
| 2023.11 | 2023-11 | Docker production bounded deployment optimization |
The project adopts a develop (development) and stable (stable) dual branch strategy. The develop branch contains the latest features but may contain breaking changes, and the stable branch has been fully tested and suitable for production deployment.
Technical advantages
- Event-driven asynchronous architecture: The event loop engine built based on Python asyncio ensures low-latency execution under high-frequency market conditions and does not block the sending of trading signals and order status updates.
- CCXT Unified Exchange Interface: Abstract exchange API differences through the CCXT library to achieve one-time development and multi-exchange deployment. There is no need to modify the strategy code to access a new exchange.
- FreqAI Adaptive Learning: An original online learning pipeline that regularly extracts features from market data and retrains the model. Built-in feature engineering pipeline and label generator lower the threshold for using ML.
- Modular strategy design: The strategy code is completely decoupled from the transaction execution logic. Users only need to implement callback functions such as
populate_indicators()andpopulate_buy_trend()to define transaction logic. - Comprehensive risk control system: A multi-layered risk control structure from the overall situation (daily trading limit, maximum drawdown) to the position level (stop loss, take profit, trailing stop), covering the main risk points of real transactions.
- Scientific citation and reproducibility: JOSS academic papers are included, backtest data can be exported, strategy parameters can be versioned and managed, and scientific research-level experiment reproducibility is supported.
How to use
Freqtrade provides multiple deployment methods:
| Deployment method | Applicable scenarios | Operation complexity |
|---|---|---|
| Docker (recommended) | Production context running | Low |
| Source code installation (pip) | Development, debugging, secondary development | Medium |
| Cloud instance + Docker | 7×24 automated trading | Low |
Quick Start Steps (Docker Solution):
- Create user data directory:
freqtrade create-userdir --userdir user_data - Create a configuration file:
freqtrade new-config --config user_data/config.json - Download historical data:
freqtrade download-data --exchange binance --days 30 - Run backtest:
freqtrade backtesting --strategy SampleStrategy - Start the real trade (dry run test first):
freqtrade trade --config user_data/config.json
It is recommended that first-time users must first run in Dry-Run mode and confirm the strategy logic and profit expectations before investing real money.
Product Pricing
| Billing dimensions | Description |
|---|---|
| Freqtrade software | Completely free (GPL-3.0 open source license) |
| Server fees | Users are responsible for their own responsibilities. Recommended cloud instances are $10~$30/month |
| Exchange handling fees | Users are responsible for responsibilities and vary by exchange and VIP level |
| FreqAI | Free built-in, no additional charge |
| Third-party data sources | Optional payment (if higher quality historical data is required) |
| Community Support | Discord Free Community Support |
| Commercial Support | No official commercial support channel |
Freqtrade itself does not have any hidden fees or paid function locks, and all core functions (backtesting, live FreqAI, Web UI) are free to use.
Application scenarios
- Personal Quantitative Trading: Individual investors use Freqtrade to deploy automated strategies to achieve 24/7 uninterrupted trading and avoid emotional decision-making. Reduce the cost of real-time trial and error through backtesting and screening strategies.
- DeFi Market Making and Arbitrage: Supports DEX/CEX exchanges such as Hyperliquid and Gate, and can be used in DeFi scenarios such as liquidity mining cooperation and cross-exchange spread arbitrage. FreqAI identifies short-term price anomalies and executes them quickly.
- Quantitative strategy research and development: The research team uses backtesting engines and hyperparameter optimization tools to quickly iterate strategy hypotheses. The FreqAI module supports feature engineering and model experiments to accelerate the transformation from research to actual trading.
- Education and Academic Research: JOSS inclusion makes it an ideal platform for quantitative trading teaching and academic research. Students can learn the complete strategy development to real market deployment process through Freqtrade.
- Asset Management: Small asset management teams can use Freqtrade as a strategy execution gateway, combining Telegram notifications and Web UI to achieve unified asset management of multiple strategies and multiple exchanges.
Applicable people
| Crowd | Adaptation value | Prerequisites |
|---|---|---|
| Personal cryptocurrency trader | Programming-free foundation configurable through Web UI, rich strategy templates | Understand basic trading concepts |
| Python developer | Can deeply customize strategies and access FreqAI for ML strategy development | Python basics and trading experience |
| Quantitative researcher | Backtesting + hyperparameter optimization + FreqAI full link | Quantitative trading knowledge |
| Small quantitative team | Low-cost strategy execution and asset management solutions | Server operation and maintenance capabilities |
| Educators/Students | Rich teaching cases and complete community resources | No special requirements |
Not suitable for the crowd: Users who have absolutely zero knowledge of programming (although the Web UI can operate basic functions, policy writing still requires some Python knowledge); enterprise users who need official SLA guarantees (the Freqtrade community provides self-service support, and there is no commercial technical support agreement).
Summary and Outlook
Freqtrade has become one of the most popular solutions in the field of cryptocurrency quantitative trading due to its advantages of being open source and free, with an active community and comprehensive functions. An ecosystem of 52,000 GitHub Stars, 340+ contributors, and 130+ enterprise users proves its maturity and trustworthiness.
Core Advantages: Completely free FreqAI machine learning integration, 15+ exchange support, backtesting + optimization + real offer full link coverage, active community ecology.
Current limitations: No official commercial support channels, advanced strategies require Python programming capabilities, high-frequency trading scenarios are limited by Python performance, and rely on exchange API stability.
Follow-up observation points: The continuous evolution of FreqAI's reinforcement learning capabilities, the further enrichment of Web UI functions supported by more DEX and Layer2 exchanges, and whether to launch a hosted version of SaaS service.
Version Info
- Freqtrade 2026.6 ：Added support for Bybit EU, optimized the FreqAI model saving and playback mechanism, and upgraded dependent libraries to Python 3.11+.
- Freqtrade 2026.5 ：Introducing a new hyperparameter optimization engine and improving backtest performance analysis reports.
- Freqtrade 2026.4 ：Reconstructed the Web UI interface and added a real-time position monitoring dashboard.
- Freqtrade 2025.12 ：Supports Hyperliquid DEX and introduces multi-time frame strategy analysis.
User Reviews