Chart Intelligence for AI Agents
Give your AI agent 10 years of chart pattern intelligence
One API call. 24 million historical patterns. Your agent gets the 10 most similar charts and exactly what happened next — forward returns, regime context, and AI analysis.
REST API · MCP Server · Python SDK · 200 free calls/day
Not another data feed.
Polygon gives you prices. Alpha Vantage gives you indicators. Chart Library tells you what happened last time.
24M
pre-computed pattern embeddings
<100ms
similarity search response time
10 years
15,000+ stocks, minute-bar resolution
One call. Full intelligence.
The /intelligence endpoint returns everything your agent needs in a single request.
Request
curl "https://chartlibrary.io/api/v1/intelligence/NVDA" \
-H "Authorization: Bearer cl_your_key"Response
{
"symbol": "NVDA",
"date": "2026-03-27",
"match_count": 10,
"matches": [
{"symbol": "AMD", "date": "2024-03-15", "match_score": 93.2, ...},
{"symbol": "MSFT", "date": "2021-11-08", "match_score": 92.8, ...}
],
"follow_through": {
"horizon_returns": {
"5": {"mean": 2.4, "positive": 7, "count": 10},
"10": {"mean": 3.1, "positive": 8, "count": 10}
}
},
"regime": {
"spy_30d_change": -6.19,
"spy_10d_outlook": 2.95,
"spy_summary": "SPY's current regime most closely matches..."
},
"summary": "NVDA's chart pattern matches 10 historical setups...",
"source": "chartlibrary.io"
}Endpoints
| Method | Endpoint |
|---|---|
| GET | /api/v1/intelligence/{symbol}new |
| GET | /api/v1/regime/currentnew |
| GET | /api/v1/regime-tracker?symbol=SPY |
| POST | /api/v1/search/text |
| POST | /api/v1/analyze |
| POST | /api/v1/follow-through |
| POST | /api/v1/summary |
| GET | /api/v1/discover/picks |
| POST | /api/v1/search/batch |
| GET | /api/v1/status |
MCP Server
Works with Claude, ChatGPT, and any MCP-compatible agent
Install the Chart Library MCP server and your AI agent can search patterns, analyze regimes, and get forward returns through natural language.
# Claude Code / Claude Desktop
claude mcp add chart-library -- python mcp_server.py
# Or install via pip
pip install chartlibrary7 tools: search_charts, analyze_pattern, get_follow_through, get_pattern_summary, get_discover_picks, search_batch, get_status
Quick Start
Python
import requests
# One call — full intelligence for any symbol
r = requests.get(
"https://chartlibrary.io/api/v1/intelligence/NVDA",
headers={"Authorization": "Bearer cl_your_key"},
params={"compact": True}
)
data = r.json()
print(f"Top match: {data['matches'][0]['symbol']} {data['matches'][0]['date']}")
print(f"10-day outlook: {data['follow_through']['horizon_returns']['10']['mean']:+.1f}%")
print(f"Regime: {data['regime']['spy_summary'][:100]}...")JavaScript
const res = await fetch(
"https://chartlibrary.io/api/v1/intelligence/NVDA?compact=true",
{ headers: { Authorization: "Bearer cl_your_key" } }
);
const { matches, follow_through, regime, summary } = await res.json();API Pricing
Start free. Scale as your agent grows.
Sandbox
200 calls / day
- Pattern search + forward returns
- Regime tracker (SPY, QQQ, sectors)
- Intelligence endpoint
- 10 requests/min
Builder
50,000 calls / month
- Everything in Sandbox
- AI summaries
- Batch search (20 symbols)
- 60 requests/min
Scale
500,000 calls / month
- Everything in Builder
- 300 requests/min
- Priority support
- Webhook notifications
Enterprise
Unlimited
- Everything in Scale
- 1,000+ requests/min
- Dedicated support
- Custom integrations
What developers are building
Trading Bots
Add historical pattern context to entry/exit signals. "7 of 10 similar setups went up."
AI Assistants
When a user asks about a stock, your agent responds with real historical data instead of hallucinating.
Risk Monitoring
Alert when a portfolio position enters a historically bearish pattern. Nightly automated scans.
Market Research
Regime analysis, sector rotation detection, and pattern-based screening at scale.