tickerflow

One await fetch() call. Crypto, stocks, forex. Zero API keys. Automatic provider fallback.

---

from tickerflow import fetch

candles = await fetch("BTCUSDT")   # crypto  → Binance → CoinGecko → Kraken → KuCoin → yfinance
candles = await fetch("AAPL")      # stocks  → yfinance → Tiingo → Finnhub
candles = await fetch("EURUSD")    # forex   → yfinance → Finnhub

One function. Any ticker. Any asset class. If the first provider fails, the next one picks up — automatically. No API keys needed for basic use.

Install

pip install tickerflow

Extras for power users:

pip install "tickerflow[pandas]"   # DataFrame output
pip install "tickerflow[serve]"    # self-hosted API server
pip install "tickerflow[chart]"    # terminal charts
pip install "tickerflow[ai]"       # Pydantic AI integration
pip install "tickerflow[all]"      # everything

Quick Start

import asyncio
from tickerflow import fetch

async def main():
    # Fetch 10 daily candles — provider is auto-selected
    candles = await fetch("BTCUSDT", interval="1d", limit=10)
    for c in candles[-3:]:
        print(f"{c.time}  O:{c.open:.2f}  H:{c.high:.2f}  L:{c.low:.2f}  C:{c.close:.2f}")

    # Want a DataFrame instead? One flag.
    df = await fetch("AAPL", interval="1d", limit=30, as_dataframe=True)
    print(df.describe())

asyncio.run(main())

CLI

# Auto-routes to the best provider
tickerflow fetch BTCUSDT 1d 10

# Force a provider
tickerflow fetch EURUSD 1d 20 --provider yfinance

# Output formats
tickerflow fetch AAPL 1d 10 --csv
tickerflow fetch AAPL 1d 10 --json

# Terminal chart — screenshot-worthy
tickerflow chart BTCUSDT 1d 30

# Self-hosted market data API
tickerflow serve --port 8000

Terminal Chart

Render OHLCV data directly in your terminal with tickerflow chart:

tickerflow chart BTCUSDT 1d 30

Renders Unicode candlestick charts with ANSI green/red coloring and a volume pane — pure stdlib, zero dependencies. Supports --no-color and respects NO_COLOR.

No extra install needed — included in the base package.

Self-Hosted API

One command starts a FastAPI server with live market data endpoints:

tickerflow serve --port 8000

Then:

curl localhost:8000/ohlcv/BTCUSDT?interval=1d&limit=5
curl localhost:8000/providers/BTCUSDT
curl localhost:8000/health

Returns structured JSON — plug it into any frontend, dashboard, or automation. No rate limit, no vendor lock-in, runs on your machine.

Requires: pip install "tickerflow[serve]"

DataFrame Output

90% of Python finance users want a DataFrame. One flag:

df = await fetch("BTCUSDT", interval="1d", limit=30, as_dataframe=True)
print(df.describe())

Returns a pandas.DataFrame with columns time (UTC datetime), open, high, low, close, volume.

Requires: pip install "tickerflow[pandas]"

Provider Routing

The router automatically detects asset class from the symbol and tries providers in order. If one fails, the next picks up.

BTCUSDT  →  Binance  →  CoinGecko  →  Kraken  →  KuCoin  →  yfinance
AAPL     →  yfinance  →  Tiingo (daily/weekly)  →  Finnhub
WM.TO    →  yfinance  →  Finnhub
EURUSD   →  yfinance  →  Finnhub

Provider	Latency	Key required	Asset classes
Binance	~50–150 ms	No	Crypto
KuCoin	~100–200 ms	No	Crypto
Kraken	~100–250 ms	No	Crypto
CoinGecko	~200–400 ms	No	Crypto
yfinance	~300–800 ms	No	Stocks, Forex, Crypto
Tiingo	~200–400 ms	Yes (free)	Stocks, ETFs
Finnhub	~200–500 ms	Yes (paid)	Stocks, Forex

Supported Intervals

1m · 5m · 15m · 1h · 4h · 1d · 1w

Caching

Responses are cached in memory with interval-aware TTLs:

Interval	TTL
1m	30s
5m	2 min
15m	5 min
1h	30 min
4h	2h
1d	4h
1w	24h

Disable: OHLCV_CACHE_ENABLED=false

Pydantic AI Integration

Use tickerflow as a tool in your Pydantic AI agents:

from pydantic_ai import Agent
from tickerflow import fetch

agent = Agent("openai:gpt-4o", instructions="You are a market analyst.")

@agent.tool
async def fetch_market_data(ctx, symbol: str, interval: str = "1d", limit: int = 30) -> str:
    candles = await fetch(symbol, interval=interval, limit=limit)
    if not candles:
        return f"No data for {symbol}"
    last = candles[-1]
    return f"{symbol}: O={last.open:.2f} H={last.high:.2f} L={last.low:.2f} C={last.close:.2f}"

See examples/pydantic_ai_agent.py for a complete working agent.

Requires: pip install "tickerflow[ai]"

Why This Exists

I built tickerflow because I was tired of wiring up different API clients for crypto vs stocks vs forex. I wanted one async fetch() call that handles everything — and falls back automatically when a provider is down.

This is the market data layer for my Pydantic AI Pine Script Expert agent. If you're building AI agents, trading bots, or data pipelines that need multi-asset OHLCV data with zero configuration — this is it.

Examples

See examples/ for runnable scripts:

• basic_fetch.py — fetch candles for crypto, stock, and forex
• multi_provider.py — inspect provider chains and observe fallback
• pandas_output.py — DataFrame output for quick analysis
• pydantic_ai_agent.py — AI agent with market data tools

Roadmap

Shipped in v0.2

• pandas DataFrame output
• tickerflow serve — self-hosted FastAPI market data API
• tickerflow chart — terminal candlestick charts
• Pydantic AI integration example

Planned

• OKX and Bybit providers
• WebSocket streaming
• Async context manager support

Contributing

See CONTRIBUTING.md for development setup, how to add providers, and PR guidelines.

License

MIT