AIStock gathers market data, news and calculations into one website. You ask about a stock, read the evidence and the risks it found, and compare how different investing styles would see it.
It can also screen for ideas against strategies you choose and test a strategy with simulated money, so nothing real is at risk. On the trade simulation page, Jev can act as the decision engine: it is shown the dated market prices, the stocks in play and the pretend account, and answers buy, sell or hold with a probability and a confidence. A language model still writes the research.
It is a Python server with a React web app, and it can run in Docker. You add a TypeSafe key only if you want the Jev option, and keys stay in the server settings, never in a saved strategy.
What you get
- Full Python and TypeScript source code under the MIT license
- The research website, its API server and the trade simulation
- Docker files, strategy templates and the test suites
Jev
Python
OpenRouter
CodeRabbit Top10
TypeScript
React
FastAPI
Convex