Social listening API: how to build one stack across six platforms
A social listening API stack is a small set of data APIs that do two jobs on each platform: find new mentions of your brand, and follow the accounts, posts and threads you already know about. One key can cover Reddit, TikTok, YouTube and news for both jobs, plus X posts and Instagram profiles.
Short answer: If no one writes code, buy a listening suite: Awario includes API access from $49 a month. With a developer, one data API key covers Reddit, TikTok, YouTube, news and X posts (Serpent API, from $0.01 per 1,000 calls — pay as you go, no subscription; pricing). For X-heavy volume, SocialData ($0.20 per 1,000 tweets) is the cheapest published X search rate I found.
What does a social listening API have to do?
It has to answer two different questions, and most buying mistakes come from treating them as one. The first is discovery: who mentioned us, our product or our competitor since the last check? The second is tracking: what happened to the accounts, posts and threads we already care about?
Discovery is a search problem. You send a keyword, a hashtag, a community name or a sound, and you get back items you have never seen. Tracking is a lookup problem. You already hold an id or a handle, and you want fresh counts, new replies or a new post from that account.
The two jobs have different costs. Discovery is open-ended, because a popular keyword can return thousands of items a day. Tracking is predictable, because you decide how many accounts and posts to follow and how often. A good stack lets you scale each one on its own.
Around those two jobs sit three plain engineering tasks: store every item under a stable id, drop duplicates, and alert only on changes that matter. None of that needs a vendor. It needs clean ids and honest empty values, which is where APIs differ most.
Suite, official APIs or data APIs: which route fits?
There are three ways to get social data into your own systems. Each one is right for somebody, so the honest answer depends on who will maintain it.
A listening suite with an API attached
Suites such as Brand24, Awario, Brandwatch and Meltwater collect mentions for you, show them in a dashboard and let you export them. You pay per month for a mention allowance, not per call. Awario lists API access on all three plans, from $49 a month for 30,000 new mentions. Brand24 starts at $249 a month for 2,000 mentions and lists the API at an additional fee of $149.
A suite is the fastest route for a marketing team. The trade-off is control: you get the sources, the mention matching and the update cadence the suite chose, and the allowance resets each month whether you used it or not.
Each platform’s official API
X, Reddit, YouTube, TikTok and Meta each publish an API. These are the most authoritative sources, and some are free within a quota. They are also five different sign-ups, five sets of terms and five auth flows. Several are gated: TikTok’s Research API is for qualifying non-commercial researchers, and Reddit asks builders to request access before using data through its API.
Per-platform data APIs under one key
Data APIs return public platform data as JSON and bill per call, per result or per credit. Examples include Xpoz, Data365, SocialData, Apify actors and Serpent API. This route suits a developer who wants listening inside their own product or warehouse, with one bill and their own matching logic.
The rest of this guide assumes you are on the third route, or mixing it with the second.
What do the official APIs allow in 2026?
Two of the five changed in ways that most listening guides have not caught up with. Here is what each platform’s own documentation said when I read it on 20 November 2026.
X
X bills new developers per resource. Its pricing page lists reading a post at $0.005 and reading a user at $0.010, and caps pay-per-usage plans at 3 million post reads per monthly billing cycle; above that you need an Enterprise plan. Recent search covers posts from the last 7 days and is open to all developers. Full-archive search back to 2006 is for pay-per-use and Enterprise customers.
That makes X’s own search a sound discovery source for X keywords, at $5.00 per 1,000 posts read. Our X API comparison prices the alternatives record by record.
Reddit’s Data API wiki sets the free tier at 100 queries per minute per OAuth client id, averaged over a 10-minute window. Its Responsible Builder Policy says you must request access and get explicit approval before accessing Reddit data through the API. Reddit publishes no self-serve commercial price; the $0.24 per 1,000 calls figure quoted across the web dates from 2023 and I could not find it on a current Reddit page. See the Reddit API comparison for third-party rates.
YouTube
This is the change most guides miss. Google’s quota page now says a project gets a default allocation of 100 search.list calls, 100 videos.insert calls, and 10,000 units per day combined for all other endpoints, reset at midnight Pacific Time. Pages that still say “a search costs 100 of your 10,000 units” describe the old pool. Either way, 100 keyword searches a day is the default ceiling for discovery on YouTube’s own API.
TikTok
TikTok’s Research API is free, but it is for academic institutions and not-for-profit research bodies in a list of regions, and applicants must show independence from commercial interests. TikTok says to expect an answer within about four weeks. A brand team is not eligible, which is why commercial TikTok listening runs through data APIs.
Meta’s Graph API offers hashtag search with a hard cap: a maximum of 30 unique hashtags within a 7-day period, called from a professional account connected to a Facebook Page. Business Discovery returns basic metadata and metrics about other professional accounts, and does not return age-gated accounts. Those are the official discovery and tracking routes on Instagram.
News
News has no single platform owner, so there is no “official” equivalent. Listening teams use a news search API that returns headline results with source and time, or a suite’s media module. Our guide to PR and media monitoring covers that workflow end to end.
How do you assemble the stack, platform by platform?
Start from the job, then pick the call. The figure below maps each platform to the call that finds new mentions and the call that follows what you already know. Every path shown is documented in our API docs.
Reddit: search, then read the threads
Use /api/reddit/search with q for keyword discovery across Reddit; each result carries an id you can pass straight to /api/reddit/comments as post_id. For communities you watch, /api/reddit/posts with sort=new returns the latest posts with score, comment count and body text. /api/reddit/subreddit gives a summary card for a community. For a ready-made recipe, see monitoring a subreddit for questions.
curl "https://apiserpent.com/api/reddit/search?q=async%20python&limit=50" \
-H "X-API-Key: sk_live_your_api_key"
TikTok: keyword, hashtag and sound
TikTok has three discovery calls. /api/tiktok/search takes a keyword in q, /api/tiktok/hashtag takes a tag, and /api/tiktok/music takes a music_id and returns videos using that sound. Each returns full video records with exact play, like, comment and share counts, plus flags that tell you whether an empty list is restricted, truncated or complete.
For tracking, /api/tiktok/profile follows a creator and /api/tiktok/video re-reads one video, optionally with include_comments. If hashtags are your main signal, read tracking a TikTok hashtag before you build alerts on one.
YouTube: search by date, then follow channels
/api/social/youtube/search takes q with order=date and published_after, which is exactly the shape of a “new videos about us since yesterday” query. /api/social/youtube/channel with include_videos=true follows a competitor or creator channel, and /api/social/youtube/video refreshes view, like and comment counts for up to 50 ids in one call.
News: one query, five engines
/api/news takes q, a freshness window such as 1d, and an engine of Google, Bing, Yahoo, DuckDuckGo or Brave. Each article carries title, source, published time, snippet and URL; Google rows also carry google_news_url. Running the same query on two engines is a cheap way to widen coverage.
X: follow known posts and accounts
/api/x/tweet reads one public post by id or url: text, author, favorites and replies, media, hashtags, links and the post it quotes. /api/x/profile reads an account by handle, with follower counts and the recent posts the profile lists; hydrate=true fills those posts with exact text and counts. Pair these with X’s recent search for keyword discovery: search finds the post ids, and the lookup keeps their counts current. The X API comparison linked above covers the search side in detail.
Instagram: follow the accounts that matter
/api/social/instagram/profile takes a username and returns followers, following, post count, bio and verification; include_posts=true adds recent posts. Use it on competitors, creators and your own accounts, and pair it with Meta’s hashtag search if you have a professional account and fewer than 30 hashtags to watch a week. Instagram profile tracking shows the snapshot method.
What does it cost per 1,000 at three volumes?
Prices below come from each vendor’s live pricing page, which I checked on 20 November 2026. Units differ, so read the second column before comparing rows. Each cell takes the cheapest monthly plan or credit plan that covers the volume and divides its price by that volume.
| Provider | Unit you pay for | At 1K a month | At 100K a month | At 1M a month |
|---|---|---|---|---|
| X API pay-per-use | Post read ($0.005 each) | $5.00 | $5.00 | $5.00 (cap: 3M post reads a month) |
| SocialData | Tweet returned ($0.0002 each) | $0.20 | $0.20 | $0.20 |
| Apify Tweet Scraper | Tweet, at least 50 per query | from $0.40 | from $0.40 | from $0.40 |
| Apify TikTok Scraper | Result | from $1.70 | from $1.70 | from $1.70 |
| Xpoz, Reddit or X query | Query (2 credits; rows free) | $16.00 (Pro plan) | $3.20 (Max + overage) | $3.92 (Max + overage) |
| Xpoz, TikTok query | Query (5 credits; rows free) | $16.00 (Pro plan) | $9.20 (Max + overage) | $9.92 (Max + overage) |
| Data365 | Credits, one network (€300 for 500,000) | Unverified | Unverified | Unverified |
| Awario | New mention (API on every plan) | $49.00 (Starter) | $1.49 (Pro) | $0.40 (Enterprise) |
| Brand24 | Mention (API +$149 fee) | $249.00 (Individual) | $6.99 (Business) | Unverified (custom) |
| Serpent API — X post lookup | Call (one post) | Default $0.10 · Growth $0.09 · Scale $0.07 · Enterprise $0.06 · Enterprise Plus $0.05, at every volume | ||
| Serpent API — X profile | Call (one account) | $0.10 on every tier ($0.15 with hydrate=true) | ||
| Serpent API — Reddit | Call (search, community); 25-record block (posts, comments) | Default $0.12 · Growth $0.108 · Scale $0.084 · Enterprise $0.071 · Enterprise Plus $0.059 | ||
| Serpent API — TikTok search, hashtag, music | Call (up to 30 videos at limit=30) | Default $3.00 · Growth $2.70 · Scale $2.10 · Enterprise $1.79 · Enterprise Plus $1.47 | ||
| Serpent API — TikTok profile, video | Call (basic) | Default $0.10 · Growth $0.09 · Scale $0.07 · Enterprise $0.06 · Enterprise Plus $0.05 | ||
| Serpent API — YouTube | Call (search, video, channel, playlist) | Default $0.20 · Growth $0.02 · Scale $0.01 · Enterprise $0.01 · Enterprise Plus $0.01 | ||
| Serpent API — News | Call (one query) | Default $0.20 · Growth $0.02 · Scale $0.01 · Enterprise $0.01 · Enterprise Plus $0.01 | ||
| Serpent API — Instagram profile | Call (one profile) | Default $0.40 · Growth $0.36 · Scale $0.28 · Enterprise $0.24 · Enterprise Plus $0.20 | ||
Xpoz charges per query, not per row: Pro is $16 a month for 7,500 credits with $3 per 1,000 extra credits, and Max is $160 for 120,000 credits with $2 per 1,000 extra. At 100,000 Reddit queries (200,000 credits) Max plus overage comes to $320, or $3.20 per 1,000 queries.
Awario’s figures use its monthly prices ($49, $149 and $399); its annual-billing prices are lower ($29, $89 and $249 a month). Data365 publishes plan prices but not the credit cost of a request, so a per-1,000 figure would be a guess.
Serpent API tiers are set by deposit size rather than by monthly volume: Growth from a $100 deposit, Scale from $500, Enterprise from $1,000 and Enterprise Plus from $1,250, with credit that does not expire. The full rules are on the pricing page.
A worked example makes the table concrete. Say a brand team runs 20 Reddit keyword searches, 10 TikTok keyword searches, 20 YouTube searches and 20 news queries a day, and re-reads 200 known X posts every day.
That is about 600 Reddit calls, 300 TikTok calls, 600 YouTube calls, 600 news calls and 6,000 X post lookups a month. On a per-call API that is a few dollars a month at most. The same 6,000 X posts read through X’s own API would be $30.00. The bill only grows large when you raise discovery depth on TikTok or pull long comment threads, so watch those two levers first.
How should you store, dedupe and alert?
Store every item under the platform’s own id, never under the keyword that found it. The same Reddit post can match three of your keywords; keyed by its id, it is one row with three tags. TikTok videos have a video_id, X posts an id, YouTube videos an id, and news articles a url.
Keep snapshots, not just the latest value. A row per item per check, with the counts at that moment, lets you alert on a change (“comments on this thread tripled in an hour”) instead of on a level. It also gives you the history a suite would charge extra for.
Treat empty values as unknown, not zero. Several of the endpoints above document this directly: a null count means the value could not be stated, and TikTok lists say whether an empty result is restricted, truncated or complete. Write those flags to your store, because an alert built on a fake zero pages someone at 3 a.m. for nothing.
Set cadence per job. Discovery on fast platforms (X, TikTok, Reddit) usually wants a check every 15 to 60 minutes for active campaigns and daily otherwise. Tracking known accounts is fine daily for most teams. News sits in between: hourly with freshness=1h during a launch or an incident, daily the rest of the time.
Finally, keep listening on social separate from listening on search results. Whether your brand page ranks, and what shows up next to it, is a different signal with different tools; see search monitoring for brand protection.
Which stack should you pick?
- Marketing team, no developer: a suite. Awario if you want an API on the entry plan; Brand24 if you prefer its dashboard and can pay the API fee. Pick by the mention allowance you will actually use.
- Developer building listening into a product: one data API key for Reddit, TikTok, YouTube, news, X posts and Instagram profiles, plus X’s pay-per-use recent search for X keywords. Serpent API covers the first part, from $0.01 per 1,000 calls — pay as you go, no subscription; X’s search covers the second at $5.00 per 1,000 posts read.
- X is most of your volume: SocialData at $0.20 per 1,000 tweets for search, or X’s own API when you need full-archive search or first-party terms.
- Credit plans suit you better than per-call bills: Xpoz, where a query costs the same however many rows it returns. Check the per-platform credit cost against your mix first; TikTok and Instagram queries use more credits.
- Academic or non-profit research: apply for TikTok’s Research API and use Reddit’s free tier and YouTube’s free quota before you pay anyone.
Whichever you choose, start small: wire one platform end to end, from call to stored row to alert, before adding the next. Most listening projects stall on the pipeline, not the data.
How we checked this
On 20 November 2026 I read the top five Google results for “social listening API” (Phyllo, Modash, Xpoz, Octolens, Syncly), the official documentation of X, YouTube and Meta, and the live pricing pages of Xpoz, Data365, SocialData, Apify (two actors), Brand24 and Awario: 17 external sources in all, 5 of them official platform pages, plus search-engine excerpts of 3 Reddit and TikTok pages. Serpent API endpoints and parameters come from our public documentation and prices from our pricing page, both read the same day. No API calls were made for this article.
What I could not verify: Reddit’s current commercial price (it is set by contract); Data365’s credit cost per request; and Brand24’s Enterprise mention allowance. TikTok’s developer pages and Reddit’s help pages would not open from my network, so facts from those two sites come from search-engine excerpts of the official pages.
Build your listening stack on one key
Reddit, TikTok, YouTube, news, X posts and Instagram profiles as JSON, from $0.01 per 1,000 calls — pay as you go, no subscription. Start with 10 free API calls.
Get Your Free API KeyExplore: Social Media API · Playground · Docs · Pricing
FAQ
What is a social listening API?
It is an API that returns public social media content and metrics, such as posts, comments, videos and profile counts, so you can find mentions of a brand or topic and follow how they develop. Some come bundled with a listening suite that collects mentions for you; others are per-platform data APIs that you call yourself and store in your own database. A useful one gives every item a stable id, exact counts and honest empty values, because those three things decide whether your alerts can be trusted.
Is there a free social listening API?
Partly: Reddit’s free tier allows 100 queries per minute per OAuth client for approved, mostly non-commercial use, and YouTube gives each project 100 search calls and 10,000 units of other calls a day. TikTok’s Research API is free for qualifying academic and not-for-profit researchers. X’s pricing page lists pay-per-use rates rather than a free tier. Among data APIs, Xpoz has 500 one-time free credits and Serpent API gives 10 free API calls, enough to test each platform before you pay.
Can I monitor Instagram hashtags through an API?
Through Meta’s Graph API, yes, within limits: hashtag search allows a maximum of 30 unique hashtags in any 7-day period, and you call it from an Instagram professional account connected to a Facebook Page. For many teams the more useful Instagram signal is account-level: follower counts and recent posts of competitors and creators, which a profile endpoint such as /api/social/instagram/profile returns by username without any account linking.
How much does social listening cost per month?
Suites charge a monthly allowance: Awario runs $49 to $399 a month and Brand24 $249 to $699 on its self-serve plans (Enterprise from $1,499) before its API fee, as listed on 20 November 2026. Per-call APIs scale with use instead. A small stack of a few thousand searches and lookups a month costs a few dollars on per-call pricing. The expensive parts are deep TikTok lists, long comment threads and X post reads through X’s own API at $5.00 per 1,000.
Should I build social listening myself or buy a tool?
Buy if the people who need the answers will read a dashboard and nobody will maintain code; a suite is faster to start and handles matching for you. Build if listening feeds your own product, warehouse or AI agent, if you need sources a suite does not cover, or if a monthly allowance wastes money in quiet months. Building means owning the pipeline: storage keyed by platform id, snapshots, deduplication and alert rules. Start with one platform and expand once that loop works.
How do I avoid counting the same mention twice?
Key every stored item by the platform’s own id rather than by the search that found it: a Reddit post id, a TikTok video_id, an X post id, a YouTube video id, or a news article URL. When a second keyword finds the same item, add a tag to the existing row instead of inserting a new one. For news, the same story often appears on several outlets; group those by headline and time window before treating each article as a separate mention.
Sources
Official platform documentation (checked 20 November 2026):
- X API pricing (pay-per-usage, 3M post-read cap)
- X API search introduction (recent and full-archive)
- YouTube Data API quota costs
- Instagram Graph API hashtag search
- Instagram Graph API Business Discovery
- Reddit Data API Wiki (search excerpt)
- Reddit Responsible Builder Policy (search excerpt)
- TikTok Research Tools: access and eligibility (search excerpt)
Vendor pricing pages (checked 20 November 2026):
- SocialData pricing
- Apify Tweet Scraper
- Apify TikTok Scraper
- Xpoz pricing
- Data365 pricing
- Brand24 prices
- Awario pricing
Pages reviewed for the gap check (vendor guides, 20 November 2026):





