StockSharp
C#/.NET algo stack with 384 connectors; source is published but no longer open source.
Last updated
What it is
StockSharp (S#) is a .NET stack for algorithmic trading: S#.API, a C# library on .NET 10; Designer, a visual strategy builder with an embedded code editor, backtester and optimiser; Hydra, a scheduled data downloader and store; Terminal, a charting and order-entry front end; and Runner, a command-line host. The applications are WPF and run on Windows only.
On 16 July 2026 the vendor replaced the repository's Apache-2.0 licence with a custom notice saying it "is not licensed under a general-purpose open source license", routing every use through an EULA that forbids modifying or redistributing the source. The connectors repository carries the same notice, and the NuGet packages point at the EULA rather than declare a licence. The code is still public; open source it is not. Development has not slowed — 5.0.238 shipped on 9 September 2026, master was last committed to on 12 September — but the newest tagged release is 5.0.0 from March 2020, so read the NuGet feed, not the Releases page.
Pricing
Free is not a trial: the library has no connector cap, and the applications run with one connector and five cloud backtests a day. Premium is $995 the first year, $498 a year after, or $2,985 once for a lifetime licence — four connectors, the Shell framework, 100 cloud runs a day. Ultimate is $1,995 the first year and $998 after, for unlimited connectors and runs plus HFT/DMA. Billing is yearly; there is no monthly plan.
Data & coverage
StockSharp sells no market data. Every quote arrives through a connector you already pay someone else for, so latency is whatever that account gives you. Hydra stores it locally and serves it back to backtests.
Integrations
384 connectors, the longest list in this category: Interactive Brokers, Schwab, Alpaca, tastytrade, TradeStation, Tradovate, Tradier, Rithmic, CQG, IQFeed, Saxo, LMAX, Oanda, cTrader, MT4/MT5, Binance, Coinbase, Bybit. The centre of gravity is not the US, though — crypto venues outnumber everything, the recent build-out is Indian, Korean and Taiwanese retail brokers, and the docs still carry a Russian family (Quik, Transaq, Plaza2, MOEX, Finam) no Western rival has. Options tooling is real: Greeks, Black-Scholes, an IV book and an option desk.
Limitations
- Not open source since July 2026; the EULA forbids modifying or redistributing the code.
- Windows-only for anything with a window.
- Python is IronPython inside Designer, not the scientific Python stack.
- Some connectors ship only from a private NuGet server, so the published list is not the installable list.
- No data of its own, no screener, no portfolio or tax reporting.
- Product pages sell "384 connectors" as free; the pricing page caps the free applications at one.
Alternatives
NinjaTrader and MultiCharts for a bought-and-supported Windows algo platform; Quantower and Sierra Chart when the front end matters more than the framework. Off .NET, the Python frameworks here cost nothing and raise no licence question.
Compare it against the rest of the backtesting and algo frameworks.
Specs
- Interfaces
- API, Python, C#, F#, Python
- Export
- CSV, JSON, Xlsx
- Asset classes
- Stocks, Options, Futures, Forex, Crypto
- Markets
- Global
- Platforms
- Desktop win, Library, CLI
- AI features
- None
- Capabilities
- Charting, Backtesting, Automation, Live trading, Paper trading, Alerts, News, Options analysis
- Pricing verified
- Capabilities verified
- Coverage verified
Also worth comparing
- Lumibot — One strategy class for backtest and live, plus a built-in LLM agent runtime.
- NautilusTrader — Rust core, Python API, one strategy that backtests and trades live unchanged.
- ib_async — The community continuation of ib_insync — same API, new maintainers, TWS still required.
- Composer — No-code rule-based stock and ETF strategies, backtested and then traded for real.
- QuantConnect — The open-source LEAN engine, plus a hosted cloud that runs it against real brokers.
- QuantRocket — Zipline and Moonshot in Docker on your own hardware, wired to Interactive Brokers.
FAQ
Is StockSharp open source?
Not any more. The repository carried Apache-2.0 from December 2015 until 16 July 2026, when a commit titled "Replace Apache license with StockSharp custom notice" swapped it for a notice stating the repository "is not licensed under a general-purpose open source license" and routing all use through the EULA. The code is still published and readable; the licence is proprietary.
What does the free plan actually give you?
The C# library with no connector cap, one connector at a time inside Designer, Terminal, Hydra and Runner, and five cloud backtest runs a day. Premium at $995 for the first year raises that to four connectors and 100 cloud runs; Ultimate at $1,995 removes both caps.
Does StockSharp run on macOS or Linux?
The applications do not. Designer, Terminal and Hydra are WPF, and the documentation marks every StockSharp GUI package Windows-only. The non-GUI NuGet packages target .NET 10.
Can I write strategies in Python?
Inside Designer, yes, but it is IronPython rather than CPython. A strategy starts with import clr and clr.AddReference("StockSharp.Algo"), so you are writing .NET in Python syntax. The docs cap it at Python 3.4 compatibility, with numpy only partly available through a .NET reimplementation and no pandas, scipy or anything else written in C.