Twitter Data Node

The Twitter Data node fetches tweets from the Twitter/X API v2. Use it to pull social sentiment, monitor specific accounts, or validate trading signals with real-time social data. It supports two modes: recent search (last 7 days) and full archive search (complete history).

Why Twitter Data Matters for Trading

Twitter/X is one of the fastest sources of market-moving information. Price action often lags social signals by minutes or hours. This node lets you tap into that signal:

• Sentiment validation — A price drop triggered your strategy, but is it real panic or just noise? Search Twitter to confirm before executing.
• Account monitoring — Track tweets from influential accounts (e.g., from:elonmusk doge) that historically move markets.
• Social volume detection — A sudden spike in tweets about a token often precedes a price move.
• News confirmation — Validate on-chain signals (funding rates, liquidations) with social context.

The strongest pattern is using this node as a mid-workflow validation step: upstream nodes detect a quantitative signal, the Twitter node confirms it with qualitative social data, and downstream nodes make the final decision.

Configuration

Search Types

Operator Reference

Most common needs are covered by the form fields (Keywords, Exclude words, From accounts, Language, the toggles). For everything else, the Custom operators field accepts any operator from the official X search operators reference ↗. Useful ones for trading workflows:

Operators are space-separated. Most are combined as AND by default; use OR and parentheses for alternation: ($BTC OR $ETH) min_faves:50.

Dynamic Queries

Toggle the f(x) button next to any text field that supports it to use a value from an upstream node. The following fields support interpolation:

• Form mode: Keywords, Exclude words, From accounts, Custom operators
• Raw mode: Query

A common validation pattern uses Keywords:

{price_data.symbol}

When the Price Data node detects a move in SOL, Keywords becomes SOL — combined with whatever filters you've set, the compiled query might be SOL -is:retweet lang:en. Language and the boolean toggles (Exclude retweets / Has media / Has links) are static per execution; everything else can be made dynamic.

Setting Up Credentials

Option 1: NickAI Credits (Recommended)

Select NickAI Credits in the credential dropdown. You're charged ~0.75 credits per tweet returned (so a 10-result search costs ~7.5 credits, a 100-result search costs ~75). No setup required.

Option 2: Your Own Bearer Token

1. Go to developer.x.com ↗ and create a project/app.
2. Generate an App-only Bearer Token (under "Keys and tokens").
3. In NickAI, go to Credentials and click Add Credential.
4. Select Twitter / X API and paste your Bearer Token.
5. In the Twitter Data node, select your credential from the dropdown.

Workflow Examples

Sentiment Validation for Trade Signals

The highest-value pattern: a quantitative signal triggers a social sentiment check before executing.

In this workflow:

1. Price Data fetches the current BTC price.
2. Conditional checks if the price dropped more than 5%.
3. Twitter Data searches for bitcoin crash OR dump -is:retweet to gauge social sentiment.
4. LLM analyzes the tweets — is it real panic or just noise?
5. Conditional routes based on the LLM's assessment.
6. Exchange buys the dip if sentiment confirms the opportunity.

Whale Account Monitor

Track tweets from influential accounts and alert when they mention specific tokens:

Funding Rate + Social Confirmation

Combine on-chain derivatives data with social sentiment for a more robust signal:

The CoinGlass node and Twitter Data node both feed the LLM. The LLM looks for confluence — negative funding rates combined with bearish social sentiment is a stronger short signal than either alone.

Pricing & Credits

When using NickAI credits, you're charged ~0.75 credits per tweet returned — the same rate across both search types (Recent and Full Archive), since X's pay-per-use API bills per post read regardless of endpoint. You only pay for tweets actually returned, not the Max Results ceiling.

Before each call, NickAI reserves the worst case (maxResults × 0.75) against your balance. After the API responds, you're charged only for the actual resultCount.

Output

After execution, the node produces the following data that downstream nodes can reference:

Example Output

{
  "tweets": [
    {
      "id": "1234567890",
      "text": "Bitcoin just broke $100k! This is massive for the entire crypto market.",
      "created_at": "2026-04-17T10:30:00.000Z",
      "author_id": "9876543210",
      "lang": "en",
      "public_metrics": {
        "retweet_count": 1542,
        "reply_count": 312,
        "like_count": 8901,
        "quote_count": 234,
        "bookmark_count": 567,
        "impression_count": 2500000
      },
      "entities": {
        "cashtags": [{ "start": 0, "end": 8, "tag": "Bitcoin" }]
      }
    }
  ],
  "includes": {
    "users": [
      {
        "id": "9876543210",
        "name": "Crypto Analyst",
        "username": "cryptoanalyst",
        "profile_image_url": "https://pbs.twimg.com/..."
      }
    ]
  },
  "metadata": {
    "source": "twitter",
    "searchType": "recent",
    "query": "bitcoin -is:retweet lang:en",
    "resultCount": 10,
    "fetchedAt": "2026-04-17T10:35:00.000Z",
    "nextToken": "b26v89c19zqg8o3fpzbke..."
  }
}

Common Errors

Next Steps

• LLM Node — Analyze tweet sentiment, extract key themes, or classify tweets.
• Conditional Node — Route decisions based on tweet count, sentiment score, or specific keywords.
• CoinGlass Node — Combine social signals with derivatives data for stronger trading signals.
• Price Data Node — Pair social sentiment with price action for divergence analysis.
• Function Node — Parse tweet data, compute sentiment scores, or aggregate metrics.