Comments and word of mouth

Comments and word of mouth runs on 13 platforms: the example run makes 3 calls and costs at most $0.0060.

Gets the comments under one post page after page, as one table: text, commenter, likes, replies, time. Every page is one call charged on its own. What people praise and complain about is for your own agent to read off the table.

Tell your AI

Install the DataXAPI skill dataxapi-comments from https://dataxapi.com/skills/dataxapi-comments.zip.

Download the packageA .zip with one SKILL.md. It runs with the dataxapi command line.SKILL.md →

Not connected yet? Connect your AI first →

Example run$0.0060at most. A failed call is never charged
Calls3with the example inputs
Platforms+913 · the same table from each
Hard ceiling10 callsno run is opened above $0.0500

Ask it like this

Any wording works. Three to start from.

What are people praising and complaining about under this Douyin video? Read three pages of comments.
Pull the comments on this post and group them: what do buyers like, and what puts them off?
Which comments under this video got the most likes, and what do they say?

What it does

Four steps. Only the second calls endpoints, so only the second costs money.

Show the plan first

Free, and needs no key: the most calls and the most money this run can come to.

dataxapi run comments platform=douyin id=7694940173372575014 pages=3 --plan

Run it, capped at that amount

Each line is one endpoint of the example run, at its listed price.

3 callsat most $0.0060A call that fails is not charged. The run stops rather than spend more than its cap.
dataxapi run comments platform=douyin id=7694940173372575014 pages=3 --yes --max-spend 0.006 --out comments.csv

Line everything up in one table

6 columns, the same on every platform; a column a platform does not give is empty. No call, no cost.

Your AI reads the table and answers

What the table means is for your AI to say. It ends with where the data came from and when.

What you get back

An illustration of the table, with the task's real columns. The names and figures in it are made up.

Your AIComments and word of mouth
sample

What are people praising and complaining about under this post?

TextCommenterLikesRepliesPosted at
这个价位能做到这样,可以了露营的老周2481362026-10-11T12:14:09Z
续航没有宣传的那么久,两天就没电山野咖啡1032582026-10-11T13:02:41Z
求链接,想给爸妈也买一个小鹿不迷路764122026-10-12T00:47:20Z

All 6 columns: comment_id Comment ID, text Text, author Commenter, likes Likes, replies Replies, published_at Posted at

3 calls · at most $0.0060Word of mouth and comment insight

Inputs it needs

Say them in your question, or let your AI ask.

platform one platform

The platform the post is on. Example: douyin

id text

ID of the post (a video, a note or a post). Example: 7694940173372575014

pages integer

How many pages of comments at most; each page is a call charged on its own. 1 to 10 Example: 3 Default: 3

The 13 endpoints it uses

One per platform and step. Each has a page with its parameters, fields and price.

Show all 13 endpoints →

Platforms it runs on

Run it on your own question.

Seeing the plan is free. A failed call is never charged.

Start free