X

X (Twitter) — Full-Archive Access

X API v2 with full tweet archive search covering the complete history of public tweets. Search, analyse, and engage with tweets, users, and conversations at scale.

Authentication

Connect via OAuth 2.0 (PKCE) to read AND write (post tweets, like, retweet, follow, bookmark). Full-archive tweet search is available to all users automatically — it uses a platform-managed token and does not require any additional setup.

Data Model

• Users: Profile lookup by username or numeric ID; user search; follower/following graphs. Fields: id (numeric string), username (handle), name (display), description (bio), public_metrics (followers, following, tweet count), url, location, verified.
• Tweets: Full-archive search; single-tweet lookup; per-user timelines, mentions, liked tweets; engagement analytics; conversation threads. Fields: text (always full — long tweets auto-promoted from note_tweet), public_metrics (likes, retweets, replies, quotes, bookmarks, impressions), entities (expanded URLs, mentions, hashtags), referenced_tweets (RT/quote/reply linkage), conversation_id (thread grouping).
• Lists: Membership for any user; list metadata, members, and list timelines.
• Bookmarks, Blocked, Muted: Private to the authenticated user (requires OAuth).

IDs and References

• All IDs are numeric strings (e.g. "1561506468"). Never pass as integers.
• Tweet URLs: https://x.com/{username}/status/{tweet_id}.
• Conversation threads share a conversation_id equal to the root tweet's ID.

Full-Archive Search

Search accesses the complete tweet archive — every public tweet ever posted, not just the last 7 days. Use start_time and end_time (ISO 8601) to scope to any historical period.

Query Operators (combine freely)

Content matching:

• keyword1 keyword2 — AND (both must appear)
• "exact phrase" — exact match
• keyword1 OR keyword2 — OR
• -keyword — NOT (exclude)

Author/mention:

• from:username — tweets by author
• to:username — replies to user
• @mention — tweets mentioning user

Content type:

• has:media / has:images / has:videos — media filters
• has:links — tweets with URLs
• has:hashtags / has:mentions

Tweet type (critical for clean results):

• -is:retweet — exclude retweets (always recommended)
• -is:reply — exclude replies
• is:verified — verified authors only
• is:quote — only quote tweets

Language:

• lang:en / lang:es / lang:ja / lang:de / lang:fr

Conversation:

• conversation_id:123 — all tweets in a thread

Geography:

• place_country:US — tweets geotagged in country

Example Queries

Research a company or product: TrueLayer -is:retweet lang:en

Find pricing/cost discussions: TrueLayer (price OR pricing OR cost OR fee OR expensive) -is:retweet lang:en

Monitor competitors: (from:stripe OR from:truelayer OR from:plaid) -is:retweet

Developer experience: "TrueLayer API" (integrate OR integration OR SDK OR developer) -is:retweet

Hiring signals: from:truelayer (hiring OR "we're looking" OR "join us") -is:retweet

Date Filtering

Use start_time and end_time parameters (ISO 8601 format):

• start_time: "2024-01-01T00:00:00Z" — from this date
• end_time: "2025-12-31T23:59:59Z" — up to this date
• Omit both for default (most recent results across full archive)

Sort Order

• recency (default) — newest first, best for monitoring
• relevancy — most relevant first, best for topic research

Engagement Metrics

Every tweet includes public_metrics:

• impression_count — total views
• like_count, retweet_count, reply_count, quote_count, bookmark_count
• Engagement rate = (likes + replies + retweets) / impressions

Research Workflows

Deep topic research:

1. Search full archive for topic with broad query and date range
2. Iterate with narrower queries based on initial findings
3. Use max_results=100 and pagination (next_token) for full coverage
4. Identify key voices by examining author profiles in results

Competitive intelligence:

1. Monitor multiple accounts with batch monitor tool
2. Search for brand mentions and sentiment across full history
3. Compare engagement metrics across competitors
4. Track narrative shifts over time with date-bounded searches

Account analysis:

1. Fetch user profile for follower/following metrics
2. Get recent tweets (exclude retweets) for content analysis
3. Run engagement analytics for performance overview
4. Review liked tweets for interest and partnership signals

Thread deep-dive:

1. Find interesting tweet via search
2. Use its conversation_id to fetch the full thread
3. Read all replies for context and community sentiment

Pagination

Paged results return next_token; pass it on the next call to get more results. Page sizes: 10-100. For thorough research, paginate through multiple pages.

Rate Limits

X enforces per-endpoint burst limits (typically 300 requests/15 minutes for search). When throttled, back off and retry after the 15-minute window resets.