postdata

Answers / Reddit

Which social media data API should an AI agent use?

Choose a social data API by the platforms and actions your agent needs, its client authentication and total workflow cost. PostToData offers read-only access to public content across 23 catalog entries (including three WeChat products) through REST and remote MCP, with OAuth in supported clients, API keys for code, source links where available and per-action prepaid pricing. It supports research, not publishing or private-account access.

Reddit · search posts · $0.003 a call · one verified website trial · updated 2026-11-13

Choose for the task you actually need

PostToData is an option for read-only research across public social platforms, including Reddit, TikTok, Xiaohongshu (RedNote), Douyin, YouTube, Facebook and Pinterest, plus Meta and Google public ad libraries. Zhihu currently supports search and question answers, not details or comments. REST and remote MCP use the same prepaid balance; supported clients can authenticate with OAuth.

Compare providers on your exact platform/action pair, response fields, client authentication and total workflow cost. A starting price is not a flat rate for every endpoint. PostToData does not provide publishing, private-account access or a guaranteed full firehose.

Hacker News search costs $0.001; Reddit and TikTok search each cost $0.003; Xiaohongshu search costs $0.03. Current prices are generated from the same catalog as billing.

Connect and check access for free

  1. In a supported Claude or ChatGPT client, add https://posttodata.com/mcp as a custom connector. Sign in with Google and approve the PostToData consent screen. Availability depends on the client, plan and workspace permissions.
  2. For Claude Code, Codex, Cursor or a custom application, follow the integration guide and supply your own API key through local configuration or an environment variable. Never paste it into a public prompt, repository or screenshot.
  3. Ask the connected client to call list_platforms and get_account. These checks are free. Confirm the account balance and the supported actions before requesting data.
  4. Review /prices.json or a REST request with dry_run=1 before executing a paid workflow. Each additional page, comment lookup or profile lookup is another call.

Run a bounded, two-platform search

One successful reddit search costs $0.003 and one successful tiktok search costs $0.003: $0.006 for this two-call workflow. Follow-up calls cost extra.

Ask: Use PostToData to search reddit and tiktok for “pour over coffee”, with limit 5 per platform. Make only those two search calls. Summarize recurring themes and link each claim to its original source. Ask me before fetching more pages, comments or profiles.

  1. After checking the price, the MCP client calls search_posts with platform="reddit", query="pour over coffee", limit=5, then the same tool with platform="tiktok".
  2. Keep the returned source URLs with the summary. Label the collection date and treat the results as a sample, not a representative ranking of the whole platform.
  3. For REST, remove dry_run=1 and send Authorization: Bearer YOUR_API_KEY. Review meta.billing after each call. Do not turn on an unbounded polling loop.
Preview prices for free; remove dry_run=1 only when ready to spend
curl "https://posttodata.com/v1/reddit/search?query=pour%20over%20coffee&limit=5&dry_run=1"
curl "https://posttodata.com/v1/tiktok/search?query=pour%20over%20coffee&limit=5&dry_run=1"

Follow a complete workflow

Limits and failures

This is read-only access to publicly visible content. Supported actions and available fields differ by platform; inspect meta.unavailable instead of assuming every metric exists. Source URLs are included when the upstream platform provides them.

Video transcripts retrieve available caption tracks. They are not speech recognition for arbitrary videos without captions. Search results and captions do not establish what appears in the video footage.

A 402 insufficient_balance response means this call was not charged: check get_account and add credit if needed. Verification or trial limits are not reset by signing in or creating another account. Upstream errors and timeouts are not charged; do not retry indefinitely.

Eligible new accounts receive $0.05 once, limited to one grant per network. Shared networks may already have claimed it. There is no recurring free credit.

Run it live

Calls run only when you press Run. Opening this page does not spend credit.

Loading…

Your results will appear here. Choose a platform and action, then press Run.

Real API calls require verification for one trial, or a signed-in account with prepaid credit.

From code

curl
curl "https://posttodata.com/v1/reddit/search?query=pour+over+coffee&limit=5" -H "Authorization: Bearer YOUR_API_KEY"
Python
import requests

r = requests.get(
    "https://posttodata.com/v1/reddit/search",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    params={"query": "pour over coffee", "limit": 5},
)
for item in r.json()["data"]["items"]:
    print(item["url"], item.get("text") or item.get("title"))

From your AI

Claude and ChatGPT clients that support OAuth can sign in without pasting a key. Follow the client setup guide. The command below is the API-key option for Claude Code.

Add the MCP server once
claude mcp add --transport http posttodata https://posttodata.com/mcp --header "Authorization: Bearer YOUR_API_KEY"

Then ask

“Search Reddit and TikTok for pour-over coffee, using one search of five results on each platform. Summarize the themes with source links; ask before extra calls.”

The model calls search_posts with platform reddit.

Good to know

  • Raw platform payloads spend context on fields a model never uses. The compact and markdown formats return only what is needed to reason and cite.
  • Agents can discover the API without a key: GET /v1, /openapi.json and /llms.txt describe everything.
  • dry_run=1 lets an agent check the cost of a call before making it.

Full reference in the docs. All Reddit endpoints are on the Reddit API page.

More answers