Turn a trading idea or Pine Script v6 strategy into a documented backtest inside ChatGPT. PineScript API validates compatible code, resolves datasets and settings, runs close-only OHLC simulations, and returns performance, risk, trades, and the exact script used. It also supports conversational analysis and input optimization, while rejecting unsupported behavior instead of silently changing strategy logic.
No reviews yetBe the first to leave a review for PineScript API
Maker
📌
I built PineScript API because testing a trading idea usually involves too much context switching.
You describe the strategy in one place, write or debug Pine Script somewhere else, configure the market and backtest settings, run the test, and then move the results into another tool just to understand what happened.
I wanted to bring that entire loop into a single ChatGPT conversation.
With PineScript API, you can describe a strategy in plain English or paste an existing Pine Script v6 strategy. It can generate or validate compatible code, resolve the dataset and execution settings, run the backtest, and return performance, risk metrics, trades, and the exact script that produced the result.
You can then continue the conversation to investigate drawdowns, compare long and short performance, review individual trades, or test multiple input configurations.
One of the most important design decisions was transparency. PineScript API does not claim to support every TradingView feature. When a strategy depends on behavior outside the supported execution model, it rejects it rather than silently rewriting the logic and presenting results for a different strategy.
Every completed run also includes an auditable receipt with the script, dataset coverage, resolved settings, engine build, and request fingerprints.
I’d love to hear how you currently test Pine strategies, which workflows feel the most repetitive, and what you’d like to see supported next.
Thanks for checking it out!
Report
Finally got my messy v6 strategy backtested without ChatGPT rewriting half my code. Loved that it just rejected the bits it couldn't handle instead of silently swapping in its own logic.
Report
the fact that it rejects unsupported behavior instead of silently tweaking your strategy logic is honestly such a respectable call, most tools would just silently rewrite your code and you'd never know
Finally got my messy v6 strategy backtested without ChatGPT rewriting half my code. Loved that it just rejected the bits it couldn't handle instead of silently swapping in its own logic.
the fact that it rejects unsupported behavior instead of silently tweaking your strategy logic is honestly such a respectable call, most tools would just silently rewrite your code and you'd never know