Compare
Chart Library vs Finnhub — the data feed versus what the data means.
These aren’t competitors. Finnhub is a solid market-data API — real-time and historical quotes, fundamentals, company news, earnings, and some alternative data, over a clean REST/WebSocket interface. If you need the raw inputs, it’s a good choice.
Chart Library sits one layer up. It doesn’t resell feeds; it answers the question those feeds can’t: “given this exact setup, what did its closest historical analogs do over the next 1/5/10 days, and how confident can I be?” — as a calibrated outcome distribution with an audit receipt, built for an AI agent to call. Most production stacks use a data API and a reasoning layer like this.
Different layers of the stack
- Finnhub = inputs. Quotes, fundamentals, news, earnings calendars — the raw numbers your system ingests.
- Chart Library = interpretation. For a specific setup, the cohort of shape-similar historical analogs, their forward-return distribution, the drivers that separated winners from losers, and a calibrated confidence band.
- You still need both. A screener or feed tells you what is; Chart Library tells you what historically followed — and unlike a raw feed, it ships the calibration: nominal 80% band held 80.8% across 300K+ audited cases.
- Agent-native. Hosted MCP (free, keyless for the core loop), REST, and SDKs — so an agent calls it the same way it calls Finnhub, but gets back a reasoned distribution, not rows.
When to reach for which
- Need live quotes, fundamentals, or a news feed → Finnhub (or your existing data vendor).
- Need “what happens after setups like this, with honest odds” for an agent to reason on → Chart Library.
- Building an autonomous trading agent → use a data API for inputs and Chart Library for the calibrated base rates it cites in its decisions.
Frequently asked questions
- Is Chart Library a Finnhub alternative?
- Not directly — it's a different layer. Finnhub serves raw market data; Chart Library is the calibrated cohort-intelligence layer that interprets a setup into a forward-return distribution. Many stacks use a data API like Finnhub for inputs and Chart Library for the outcome read.
- Does Chart Library provide real-time quotes or fundamentals?
- No. It's a retrieval-and-calibration engine over historical analogs, not a market-data feed. Pair it with a data vendor (Finnhub, Polygon, etc.) for live inputs; Chart Library supplies the calibrated historical base rates.
- Does it predict prices?
- Never. It returns what historical analogs of a setup did next, as a calibrated distribution with a coverage receipt — similarity-only by design, no directional forecast.
Try it
Run a cohort_analyze call.
Free Sandbox tier — 1,000 calls/day, no authentication. MCP install for Claude or Cursor takes 30 seconds.