Fetch social data without a browser

A profile, as HTTP

A TikTok profile is GET /v1/tiktok/profile with handle. The recorded call:

1curl -G "https://api.socialscrape.dev/v1/tiktok/profile" \2  -H "x-api-key: $SOCIALSCRAPE_API_KEY" \3  --data-urlencode "handle=khaby.lame"

1import os2import requests34params = {5    "handle": "khaby.lame",6}7res = requests.get(8    "https://api.socialscrape.dev/v1/tiktok/profile",9    headers={"x-api-key": os.environ["SOCIALSCRAPE_API_KEY"]},10    params=params,11)12body = res.json()1314if body["ok"]:15    print(body["data"])16else:17    print(body["error"]["code"], body["error"]["message"])

1const url = new URL("https://api.socialscrape.dev/v1/tiktok/profile");2url.search = new URLSearchParams({3  "handle": "khaby.lame"4}).toString();5const res = await fetch(url, {6  headers: { "x-api-key": process.env.SOCIALSCRAPE_API_KEY! },7});8const body = await res.json();910if (body.ok) {11  console.log(body.data);12} else {13  console.error(body.error.code, body.error.message);14}

An Instagram profile is GET /v1/instagram/profile with handle. The recorded call uses nasa.

1curl -G "https://api.socialscrape.dev/v1/instagram/profile" \2  -H "x-api-key: $SOCIALSCRAPE_API_KEY" \3  --data-urlencode "handle=nasa"

1import os2import requests34params = {5    "handle": "nasa",6}7res = requests.get(8    "https://api.socialscrape.dev/v1/instagram/profile",9    headers={"x-api-key": os.environ["SOCIALSCRAPE_API_KEY"]},10    params=params,11)12body = res.json()1314if body["ok"]:15    print(body["data"])16else:17    print(body["error"]["code"], body["error"]["message"])

1const url = new URL("https://api.socialscrape.dev/v1/instagram/profile");2url.search = new URLSearchParams({3  "handle": "nasa"4}).toString();5const res = await fetch(url, {6  headers: { "x-api-key": process.env.SOCIALSCRAPE_API_KEY! },7});8const body = await res.json();910if (body.ok) {11  console.log(body.data);12} else {13  console.error(body.error.code, body.error.message);14}

Both return a profile: handle, name, bio, avatar_url, verified, is_private, external_url, links and stats (followers, following, posts, likes, views). Nulls are values the platform did not send.

Instagram's profile leaves stats.posts null on purpose. The reference says to use GET /v1/instagram/profile/post-count for that number. Do not treat the null as zero posts.

You did not set a user agent, a cookie jar or a puzzle token. The request is the header and the query. That is the whole client.

The limit you still have

You stop hitting the website's browser checks. Packs and plans have no rate limits here, and free accounts have a lower limit. It is per team, not per key, so more keys do not raise it. The rate limits page has the detail.

When a free account passes its limit the status is 429, the code is rate_limited, retryable is true, and credits_charged is 0. The retry-after header is the number of seconds to wait. Wait, then send the same request. Refused requests are free.

That is a different limit from Instagram's. Calling this API does not tell you Instagram's current threshold, and it does not raise it. Your code should branch on error.code, not on a captcha page you will not receive.

Cache, so you call less

max_age_days from 0 to 30 asks for a cached copy. A hit costs 0 credits, sets cached true, and sets cached_at to when the data was first fetched. These profile endpoints keep results 30 days.

Nothing is served from the cache unless you ask. The cache key is the endpoint plus its parameters. trim and raw do not split the key. A different handle does.

If you poll the same profile every minute, send max_age_days and you will not spend a credit on a fresh fetch each time. A miss still fetches and still costs a credit. See caching.

Failed requests cost 0 credits as well. not_found, invalid_params and unsupported_url are failures. Stop retrying them. upstream_blocked and rate_limited are in the errors list. Read retryable on the one you got.

What stays impossible on purpose

Gaps are documented per endpoint, and they are not a puzzle to bypass.

Instagram comments return about 14 comments and has_more false. Replies are not available. include_replies is ignored. TikTok profile videos return about 10 videos and has_more false, with likes null, because the embed page does not have them.

LinkedIn company posts stop at one block, up to 10, and the cursor stays null. A private account stays private.

region on some routes is a country code for results that vary by country. On TikTok search the reference calls it a region hint, not a proxy country. There is no parameter for a residential proxy, a TLS fingerprint or a solved captcha token.

Do not add one. The request will reject unknown parameters as invalid_params.

Batching does not dodge the team limit. A batch of up to 100 requests runs on our side and saves round trips. Each item still counts toward the minute allowance.