Pionex.AI does not hand you a fixed library of copy-and-paste strategy files. Instead you describe the strategy you want in plain language and a dedicated agent writes the code, backtests it, and — if you choose — hosts it to emit live signals. Three parts of Pionex.AI do this work: the Quant agent (build and test strategies), the Coder agent (write and fix code and indicators), and the Agents Market (add ready-made specialist agents or build your own). This guide covers all three.
Is there a ready-made strategy template I can just copy?
There is no gallery of pre-written strategy files to download. What is "ready-made" on Pionex.AI is:
The pre-built agents in the Agents Market (see the roster below) — each already primed with a trading approach.
The Quant workspace quick-start actions — Write Strategy, Run Backtest, Write Indicator, and Write Agent — which start you off without a blank page.
To get a working strategy you tell an agent what you want (the asset, the timeframe, and the idea), and it generates the code for you. Think of it as briefing a quant developer, not picking a template off a shelf.
Writing a strategy with the Quant agent
The Quant tab opens a workspace built for strategies, with a live chart beside the chat and quick-start actions instead of a single chat box. To create a strategy, describe it in the chat or start from Write Strategy. The clearer your brief, the better the result — tell it:
The asset — e.g. BTC, ETH (crypto pairs).
The timeframe — daily, 4-hour, or 1-hour.
The idea — an indicator or signal strategy (for example EMA/SMA cross, RSI, Bollinger, MACD, ATR, or a breakout), or a grid / market-making strategy.
If your request is vague, the agent asks a couple of clarifying questions first, then writes the Python strategy and runs a backtest. A result card appears in about one to two minutes with the metrics and a trade-by-trade breakdown; ask the agent to interpret or refine it, and it will iterate — adding a stop-loss, a filter, or longer periods — and re-run.
A few things worth knowing about backtests:
Supported intervals are 1-hour, 4-hour, and 1-day, and a single run covers up to 1,000 bars. For a long period the agent picks a coarser interval so the whole window fits rather than silently cutting it short.
A backtest is a historical simulation on real market data — it is not a live order and moves no money.
Each run draws on your plan's monthly backtest allowance (see Plans below).
Going live (hosting)
When you are happy with a strategy you can ask the Quant agent to host it. A hosted strategy keeps running on new bars and pushes buy/sell signals into the chat so you can act on them. Hosting produces signals for you to review and act on — it does not place, cancel, or manage orders on your account by itself.
Using the Coder agent
The Coder agent is a general programming assistant. Use it to:
Explain, write, debug, refactor, or review code in any major language.
Write crypto quantitative trading strategies, indicators, and bots — including TradingView Pine Script and Python (ccxt, backtrader, pandas-ta) and other quant frameworks.
While it codes, Coder can pull real crypto market data — candlesticks, current price and 24h stats, available trading pairs, live order-book depth, and recent trades — to test logic and build examples, and it can search the web for API docs and references. These market-data tools cover cryptocurrency pairs only, not stocks or forex.
What Coder does not do: it does not read your Pionex account data, place or manage trades, or run backtests. To run a strategy against historical data, use the Quant agent; for account questions use Support; for markets and your portfolio use Pai.
Coder vs. Quant, in one line: Coder writes code and indicators; Quant writes and runs full strategy backtests and can host them for live signals. For building and testing a trading strategy, start with Quant; reach for Coder when you want custom code, a custom indicator, or help debugging your own script.
The Agents Market: ready-made agents
Click + Add Agent at the bottom of the sidebar to open the Agents Market. Adding agents is free — there is no per-agent price; what you pay for is your Pionex.AI plan. The market has two parts, plus the option to build your own.
Official Agents (built and maintained by Pionex)
Coder — writes and fixes code, indicators, and trading-strategy scripts.
Quant — builds Python strategies, runs backtests, and can host them for live signals.
Support — account, KYC, deposits and withdrawals, fees, bots, and connecting you to a human.
(Pai, your personal finance and markets agent, is available on every plan alongside these.)
Marketplace (community specialist agents)
Pavan Trading Guru — a trading-focused assistant for market analysis, strategy ideas, and decision support, with entry/exit and grid-bot setups.
Satoshi Assistant — a market secretary that applies regime detection and a BB-MACD approach to surface actionable entry signals and grid-bot parameters.
@doinvestfuture — a cross-market allocation analyst covering crypto, tokenized US stocks, precious metals, and Taiwan equities, strong on trend reading, leverage decisions, and risk control.
Hao Assistant — an AI-era growth-stock research assistant that teaches you to think in frameworks (stock scoring, entry timing, risk discipline) rather than making the call for you.
Click Add Agent on any card to add it to your sidebar. The exact marketplace line-up changes as new agents are added.
These community agents run on the Pai engine, so some go beyond analysis and can propose a futures Grid bot or a perp order for you — always as a confirmation card you approve, never placed on their own. They cannot open spot orders, spot Grid bots, or Dual Investment; for those, use Pai.
Build your own (Custom Agent)
Click Custom Agent in the top-right of the Agents Market, give it a name and a detailed description of what it should do, pick an avatar, and click Save to add it to your sidebar.
Plans
Coder and Quant are Pro agents: they draw on your Pionex.AI plan's monthly coding-token and backtest allowances, while Pi and Support are available on every plan. The Free plan includes 100,000 coding tokens and 3 backtests per month; paid plans (Plus, Pro, Max) raise both and unlock all models. For the full plan comparison and how billing works, see Getting Started with Pionex.AI and Pionex.AI Pricing Plans Overview.
Frequently asked questions
Q: Which agent writes a strategy — Coder or Quant?
A: Use Quant to build and backtest a full strategy (and host it for live signals). Use Coder for custom code, custom indicators, or debugging your own script. Quant is the place to start for a trading strategy.
Q: Do I need to know how to code?
A: No. Describe the asset, timeframe, and idea in plain language and the agent writes the Python for you. You can read, run, and refine it without writing code yourself.
Q: Are there pre-written strategy templates to copy?
A: Not as downloadable files. The "ready-made" options are the pre-built agents in the Agents Market and the Quant workspace quick-starts (Write Strategy, Run Backtest, Write Indicator, Write Agent). You get a strategy by briefing an agent, and it generates the code.
Q: Will a backtest or a hosted strategy trade for me?
A: No. A backtest is a historical simulation, and hosting sends buy/sell signals to your chat. Neither places live orders or moves funds — acting on a signal is always your decision.
Q: What does it cost to add an agent from the Agents Market?
A: Adding agents is free. There is no per-agent charge — you pay only for your Pionex.AI subscription plan, and Coder and Quant usage draws on that plan's coding-token and backtest allowances.