PricingOne plan$7 a month

Run always-on automated trading
strategies in the cloud

Ship your strategies from a GitHub repo to an always-on cloud container with ultra-low-latency routing to your broker. Any language, any framework, deployed in a click.

Deploy

Pro Trader

$7/ monthUSD

One always-on container, deployed from GitHub, hosted next to your broker.

  • Always-on runtime, restart-safe
  • Deploy from a GitHub repo, any language
  • Live logs and container vitals
  • MCP control surface for Claude and Cursor
  • Encrypted broker credentials and API keys
  • No shared workloads, trading-only host
Deploy

Container scales to

  • CPUUp to 1,000 vCPU
  • MemoryUp to 512 GB
  • StorageUp to 1 TB
  • RegionsEU, US-East
  • StrategiesUnlimited

Strategies hungry for more headroom? Scale CPU, memory and storage up to the container ceiling, or add a second container, all from the control panel. Billing prorates to the day.

Everything included

Nine reasons why you’ll love us

The $7 plan is not a teaser. Every container ships with the same nine capabilities, no matter how far you scale it up. You pay for capacity if your strategies outgrow it, never to unlock features.

  • 24/7 runtime

    Restart-safe containers, monitored uptime, auto-recovery on crash. Your strategy keeps running while you sleep.

  • Any language

    Python, TypeScript, Rust, Go, C++, JVM, plus 22 more. If it runs in a Linux container, it runs on TradingHost.

  • GitHub-native deploys

    GitHub is the source of truth. Every push to main rebuilds your container. Roll forward, roll back, instantly.

  • Broker-proximate regions

    EU and US-East regions placed for broker and exchange proximity. Ultra-low-latency routing tuned to your broker, not the average general-purpose cloud web request.

  • Many strategies, one container

    Run multiple strategies side by side. Per-strategy deploys do not disrupt siblings running in the same runtime.

  • MCP control surface

    Claude, Cursor and ChatGPT can deploy, monitor and roll back strategies through the native MCP server. No glue code.

  • Live observability

    Streaming logs, container vitals, deploy history, latency to broker. All in the control panel, no hooking required.

  • Encrypted secrets

    Broker credentials, API keys, exchange tokens. Stored once at rest, injected at runtime, never leave the box.

  • No noisy neighbours

    Every workload on the platform is a trading workload. No video transcoders or sleeping CRMs sharing your slice.

vs a generic VPS

Cheaper, faster, less to babysit

If you are paying for a Windows VPS to keep a desktop terminal alive, you are paying for someone else’s video stream to share your slice. Here is what changes when the host is built for trading.

Feature-by-feature comparison of a generic VPS against the TradingHost runtime.
FeatureGeneric VPSTradingHost
RuntimeNot included. Bloated Windows operating system.Included. Managed container, language-agnostic, restart-safe.
DeploysNot included. Remote desktop, drag-and-drop, hope it sticks.Included. Push to GitHub main. Container rebuilds automatically.
LocalityNot included. Generic data centre, average web latency.Included. EU and US-East regions. Routes tuned for broker endpoints.
ObservabilityNot included. RDP in and tail a log file. Hope you set up monitoring.Included. Streaming logs, vitals, deploy history. Built in.
AI workflowNot included. None. Bring your own glue code.Included. Native MCP for Claude, Cursor and ChatGPT.
TenancyNot included. Shared with video streams, batch jobs, sleeping CRMs.Included. Trading-only host. No noisy neighbours.
PriceNot included. Typically $20 to $40 per month.Included. $7 per month. Upgrade if you outgrow it.

FAQ

Questions traders actually ask

One always-on cloud container in a broker-proximate region, sized for serious automated trading and ready to scale up to 1,000 vCPU, 512 GB of memory and 1 TB of storage as your strategies grow. Deploy any-language code straight from a GitHub repo, watch live logs and container vitals from the control panel, and run as many strategies side by side as your portfolio needs. It just runs, 24/7.

Anything that runs inside a Linux container. Python, TypeScript, JavaScript, Rust, Go, C++, C#, Ruby, R, Julia, Swift, Kotlin, Scala, Haskell, OCaml, F#, Solidity, Elixir, Clojure, Erlang, Lua, PHP, Dart, Perl, Crystal, Zig, Nim and JVM languages all work out of the box. If your strategy can be containerised, the runtime can ship it.

Link your GitHub account, point TradingHost at a repo, pick a region and click Deploy. Every push to your main branch rebuilds and ships. An AI assistant connected via MCP can do the same thing from a prompt.

Yes. A container can host multiple strategies, and a strategy-level deploy will not disrupt siblings running in the same runtime. Run a portfolio of strategies on one container, then split into separate containers when you need to.

Crypto exchanges like Binance, Bybit, OKX, Coinbase and Kraken, plus anything reachable via CCXT. Forex and CFD brokers like IC Markets, Pepperstone, Exness, FP Markets, Switch Markets, OANDA and IG. Stocks and futures via Interactive Brokers. Options through partner APIs. See the full broker list for current connector status.

TradingHost ships a native MCP server, so AI assistants can authenticate, deploy strategies, inspect logs and manage your account directly. MCP is the control surface for deployment and observability, not the execution path for trades. There is a step-by-step Cursor workflow that walks the whole loop end to end.

Yes. Scale CPU, memory or storage anytime from the control panel, all the way up to the per-container ceiling. Changes apply instantly and billing prorates to the day, so you only ever pay for the capacity your strategies actually use. Need more than one box? Add a second container with a click.

Your container stops billing at the end of the current period and your account drops back to the free dashboard view.

A VPS hands you a raw operating system and a remote desktop. TradingHost hands you a managed runtime, GitHub-driven deploys, broker-proximate regions, live observability, encrypted secrets and an MCP control plane. None of that is a feature you have to bolt on yourself. The full side-by-side breakdown covers cost, migration paths, and where each one still makes sense.

Ship a strategy this weekend

One push to your main branch, one click in the control panel, and your code is running in the cloud next to your broker.

Deploy