AlphaVision Developer API

Structured market signals for humans and AI agents. Bearer-token REST API with a free SPY surface. Signal-only — no execution. Not financial advice.

Get an API key OpenAPI spec ↗

Quickstart

Free SPY (no auth)

curl https://api.alphavision.market/v1/signals/SPY

Python with API key

import os
import requests

resp = requests.get(
    "https://api.alphavision.market/v1/signals",
    headers={
        "Authorization": f"Bearer {os.environ['ALPHAVISION_API_KEY']}",
    },
    timeout=10,
)
resp.raise_for_status()

for signal in resp.json():
    print(
        signal["ticker"],
        signal["horizon"],
        signal["signal"],
        round(signal["confidence"] or 0, 3),
    )

Wire up an AI agent

# Tell your agent:

Base URL:        https://api.alphavision.market/v1
Authentication:  Authorization: Bearer <api_key>
OpenAPI spec:    https://api.alphavision.market/v1/openapi.json

# Quick rules of engagement:
# - Free SPY signal works without a key.
# - Other tickers require an active subscription.
# - Per-(key, ticker) cooldown is 10 minutes — no benefit to polling faster.
# - Signal-only. Never execute trades; surface them for human review.

Authentication

Pass a bearer token in the Authorization header:

Authorization: Bearer av_live_<your_key>

Keys are issued from your account page. One active key per member. Rotating revokes the old key immediately.

Rate limits

10-min cooldown per (key, ticker) for single-ticker endpoints.
10-min cooldown per key for the bulk /v1/signals endpoint.
Anonymous SPY = 1 request per 10 min per IP. About 6/hour.
Metadata endpoints (/me, /tickers, /status): 30/hour per caller.

AlphaVision signals update weekly to quarterly — polling faster than 10 minutes has no benefit.

Endpoints

Base URL: https://api.alphavision.market

Coverage spans five asset classes — stocks, crypto, macro ETFs, sector rotation, and volatility — roughly 80 instruments at weekly, monthly, and quarterly horizons. Volatility and sector signals report UP/DOWN rather than BUY/SELL. Call /v1/tickers to enumerate exactly what your key can read.

GET/v1/signals/{ticker}

Latest published signal for one ticker.

Returns the most recent prediction with confidence, the expected-move corridor (expected_move_pct + low/high), valid_until, next_signal_at, model version, a short track record, and the last few signals for context. Optional ?horizon=weekly|monthly|quarterly. SPY is publicly accessible.

GET/v1/signals

Latest signal for every (model, ticker) the caller can access.

Bulk endpoint — one row per (model, ticker) you can access, across every entitled asset class and horizon. Optional ?horizon=weekly|monthly|quarterly. Anonymous callers see only SPY. The bulk cooldown is independent of the per-ticker cooldown.

GET/v1/tickers

Accessible universe summary.

Returns each (ticker, horizon) the caller can read, with asset_class, model_version, and launch_date. Use this to learn what is queryable before fetching signals.

GET/v1/me

Caller identity, tier, and cooldown rules.

Inspect the resolved tier (Free / Equity / Multi-Asset / Quant), the key prefix, and the current cooldown values. Returns Free + no email for anonymous callers.

GET/v1/status

Data freshness by asset class.

Latest published signal_date for each asset class. Lets agents check liveness before consuming.

Example signal response

{
  "ticker": "SPY",
  "asset_class": "stocks",
  "horizon": "weekly",
  "signal": "BUY",
  "confidence": 0.92,
  "expected_move_pct": 1.18,
  "expected_move_low_pct": -0.94,
  "expected_move_high_pct": 3.30,
  "horizon_days": 5,
  "signal_date": "2026-05-10",
  "valid_until": "2026-05-15",
  "next_signal_at": "2026-05-16T20:00:00-04:00",
  "model_version": "Iris-v2.1",
  "launch_date": "2026-01-04",
  "track_record": { "sample_size": 419, "win_rate": 0.90 },
  "recent_signals": [
    { "signal_date": "2026-05-03", "signal": "SELL", "confidence": 0.96 }
  ],
  "disclaimer": "Informational only. Not financial advice. ..."
}

expected_move_pct is the signed expected move over the horizon (positive = up), with expected_move_low_pct / expected_move_high_pct giving the ~20th–80th percentile corridor and horizon_days the calendar span — the same magnitude estimates behind the on-site FutureChart. These are null when a signal side has no resolved history yet.

Errors

All errors share the same envelope. Use error.code for programmatic branching; message is for humans.

{
  "error": {
    "code": "ticker_not_entitled",
    "message": "'AAPL' requires an active subscription. SPY is free.",
    "docs_url": "https://alphavision.market/developers#ticker_not_entitled"
  }
}

code	HTTP	When
unauthenticated	401	Authorization header missing or empty.
invalid_key	401	Bearer token is malformed, unknown, or revoked.
subscription_inactive	403	Member's subscription is not active.
ticker_not_entitled	403	Caller's tier does not cover this ticker or asset class.
ticker_not_found	404	No active model serves the requested ticker.
rate_limited	429	Per-(key, ticker) cooldown not yet elapsed. Retry-After header indicates seconds.
invalid_request	400	Malformed parameters.
server_error	500	Unexpected error. Should not occur — report to support.

What this API does not do

It does not place trades. No execution permissions.
It does not provide personalized financial advice.
It does not recommend position sizing.
It does not include historical backtest data for arbitrary date ranges.
It does not include heatmap images or per-feature attribution.

Past performance does not guarantee future results. Use signals as inputs to your own decision process — always with a human in the loop.

OpenAPI spec Get an API key Terms Contact support