Stock Sentiment Score API: Reddit, X, News & YouTube

One sentiment number per US stock that blends social sentiment from Reddit, X, YouTube, Hacker News and Substack with the news flow, plus the 30-day baseline that tells you whether today's reading is unusual. Give it tickers, get one row each.

Who it is for

• Traders who want a market-mood read on a watchlist before the open, without scraping six sites.
• Analysts and quant researchers who need a consistent daily sentiment series to join to prices.
• AI agents that need sentiment as one structured row instead of a thousand posts to summarize.
• Developers who wanted Stocktwits API access and cannot get it. This actor carries no Stocktwits data, but it answers the same question, what the crowd is saying about a ticker and which way it leans, from Reddit, X, YouTube, Hacker News, Substack and news, as one composite Score.

What you get per ticker

• SentiSense Score: a 0-centered composite of tone and mention volume. Positive means the conversation leans bullish and is loud enough to matter. Today's reading, the 30-day average, the 30-day change, and a seven-band label from Strong Bearish to Strong Bullish.
• Direction and trend: Bullish, Neutral or Bearish, plus UP, DOWN or FLAT.
• Mention momentum: today's mention count against the 30-day daily average, and share of voice.
• Per-source tone: how News, Reddit, X, YouTube, Hacker News and Substack each lean, loudest source first, with each source's share of the mentions.
• Story drivers with their own tone, and a narrative explaining why the Score sits where it does.

Built by SentiSense, which ingests news, social media and market data all day and distills it with AI.

Data by SentiSense. API docs: sentisense.ai/docs/api.

Input

Field	Type	Default	Notes
tickers	string[]	required	US stock or ETF symbols, e.g. ["NVDA", "TSLA"]. Up to 25 per run, case-insensitive, duplicates removed.
includeDrivers	boolean	true	Include the top stories driving each ticker, each with a tone reading.

Output

One dataset row per ticker. Illustrative row, every value invented, bySource shortened:

{
  "ticker": "NVDA", "companyName": "NVIDIA Corporation", "asOf": "2025-11-03",
  "sentisenseScore": 18.7, "sentisenseScoreAvg30d": 24.1, "sentisenseScoreDelta30d": -3.6,
  "scoreLabel": "Bullish", "direction": "Bullish", "latestDirection": "Slightly Bullish", "trend": "DOWN",
  "mentions": 1420, "mentionsAvg30d": 1185.0, "socialDominance": 0.031,
  "bySource": [
    {"source": "News", "direction": "Bullish", "mentionShare": 58, "value": 0.22},
    {"source": "Reddit", "direction": "Neutral", "mentionShare": 27, "value": 0.03},
    {"source": "Hacker News", "direction": "Neutral", "mentionShare": 4, "value": 0.05}
  ],
  "drivers": [{"title": "Example headline about a product launch", "tone": 0.41}],
  "narrative": "Example narrative explaining the reading."
}

Reading the fields:

• sentisenseScore and latestDirection are today's reading; sentisenseScoreAvg30d, scoreLabel and direction describe the 30-day regime. Before today's reading lands, the two today fields are null and the regime fields still answer.
• scoreLabel bands: Strong Bearish, Bearish, Slightly Bearish, Neutral, Slightly Bullish, Bullish, Strong Bullish. direction collapses them to three.
• bySource[].value is tone on -1 to +1; mentionShare is a whole-number percent summing to 100. socialDominance is a fraction of all tracked conversation (0.031 = 3.1%), null when no signal.
• Uncovered tickers return {"ticker": "XYZ", "error": "no_coverage"}; failed requests return "error": "request_failed" or "http_<status>". Neither is charged.

Pricing

$0.05 per ticker analyzed, charged only when a ticker returns a sentiment row. No coverage or a failed request costs nothing. A 25-ticker run with 23 covered names costs $1.15.

FAQ

How fresh is the data? The Score is a daily reading, not a live feed: it is recomputed once a day and today's reading lands late morning US Eastern time (later on weekends). Before that, sentisenseScore and latestDirection are null while the 30-day fields are complete, so a pre-market run still gives a usable regime read.

Which tickers are covered? US-listed stocks and ETFs tracked by the data source. The cheapest check is to run the symbol: an uncovered one comes back as a free no_coverage row.

Does it work for ETFs? Yes. A tracked ETF such as SPY returns the same sentiment row as a stock and is charged like any other analyzed ticker.

Does it include Stocktwits? No. There is no Stocktwits data in this actor: no posts, no message stream, no bullish or bearish tags from that service, and no affiliation with it. The social side of the Score comes from Reddit, X, YouTube, Hacker News and Substack, with news alongside. If you cannot get Stocktwits API access and the question you need answered is which way the crowd leans on a ticker, this is one way to get that read as a daily number.

How much of the Score is social? It varies by ticker, and bySource[].mentionShare tells you for each one. On most large caps News carries the largest share, often two thirds to three quarters, with Reddit and X next. On retail favorites Reddit can carry most of the conversation. A small source can show a share of 0 or 1 percent on a quiet day.

What does no_coverage mean? The symbol is not tracked. It is not an outage. Check the spelling and that it is a US listing. That row is not charged.

How often should I run it? Once a day after late morning US Eastern for a fresh Score; hourly runs buy the same Score again. Mention counts and drivers can move intraday if you want those.

Is the Score the same as sentiment? No. Per-source tone is a -1 to +1 polarity. The Score is volume-aware and unbounded, so a mildly positive name with a mention surge can outscore a quietly adored one. Compare Scores to Scores and tone to tone.

Educational and informational use only. Not investment advice.