Live competitive intelligence

Competitive intelligence without the analyst or the invoice.

Get it from the live web in a single API call. Track what competitors ship, charge, and say, structured to your schema and cited to the source. Backed by Mozilla, so your research stays private, ephemeral, and never trains a model.

Start for free →Try it live →

✓Today's moves, not last week's crawl✓Diff pricing and feature changes✓Cited, ready to hand off✓Private, backed by Mozilla

research.ts

import Tabstack from '@tabstack/sdk'

const client = new Tabstack()

// Ask about a competitor. Live web, cited.
const stream = await client.agent.research({
  query: 'What did Stripe change in the last 30 days?',
  mode: 'fast',
  nocache: true,
})

for await (const event of stream) {
  if (event.event === 'complete') {
    console.log(event.data.report)
  }
  if (event.event === 'error') {
    console.error(event.data.error.message)
  }
}

Own it vs. buy it

You are renting a dashboard for data you could pull yourself.

Doing it by hand doesn't scale, and it's stale the moment you stop. Buying it means renting a pipeline, dashboard, and alerting on top of data that's already public. Both leave you a step behind. Build the piece you actually need on a handful of calls and own it. The right comparison is total cost of ownership, not the sticker price of one call.

Buy a tool

A monthly seat whether you use it or not
You pay for the pipeline, dashboard, and alerting
A pipeline you cannot inspect or verify

Monthly seat

Build on Tabstack

owned

Usage-based credits across five endpoints
Build the piece you actually need and own it
Open every source and verify the claim yourself

usage-basedyou own it

/research

Ask about a competitor, get today's answer, not last week's crawl.

Give /research a question about a competitor. It reads the live web, finds the sources, synthesizes the answer, and returns citations you can open. No index sitting between you and the truth, no orchestration layer to build. Run it fresh on every call so the answer reflects what is live right now.

Features

✓Live-web, multi-source research in one call
✓Citations and source URLs on every claim
✓nocache for always-fresh results
✓Fast and Balanced modes
✓Streams over SSE

research.ts

TypeScript

import Tabstack from '@tabstack/sdk'

const client = new Tabstack()

// Ask about a competitor. Live web, cited.
const stream = await client.agent.research({
  query: 'What did Stripe change in the last 30 days?',
  mode: 'fast',
  nocache: true,
})

for await (const event of stream) {
  if (event.event === 'complete') {
    console.log(event.data.report)
  }
  if (event.event === 'error') {
    console.error(event.data.error.message)
  }
}

extract.ts

TypeScript

import Tabstack from '@tabstack/sdk'

const client = new Tabstack()

// Full schema: github.com/Mozilla-Ocho/tabstack-schemas/blob/main/b2b-intel/competitor-profile.json
const profile = await client.extract.json({
  url: 'https://competitor.com',
  json_schema: {
    type: 'object',
    properties: {
      company_name: { type: 'string' },
      tagline: { type: 'string' },
      category: { type: 'string' },
      value_propositions: { type: 'array', items: { type: 'string' } },
      target_customers: { type: 'array', items: { type: 'string' } },
      key_features: {
        type: 'array',
        items: {
          type: 'object',
          properties: {
            name: { type: 'string' },
            description: { type: 'string' },
          },
        },
      },
    },
  },
  effort: 'standard',
})

console.log(profile) // one call, a structured competitor profile

/extract/json + /extract/markdown

Turn any competitor page into structured data you can diff.

Point /extract at a competitor's homepage, pricing page, or changelog and define the JSON shape you want back, or start from the ready-made competitor-profile schema. No parsing, no cleanup on JS-heavy pages. Run it on a schedule, store the output, and diff it over time so you catch a price change or a shipped feature the day it happens, not the month you notice.

Features

✓Ready-made schemas for company, pricing, changelog, and more
✓Schema-driven extraction, no parsing code
✓Clean markdown from any URL for change tracking
✓Run on a schedule and diff over time
✓geo_target for region-specific pricing

Managed, backed by Mozilla

Managed intelligence, backed by Mozilla, priced by usage.

The browser and the models run on Tabstack, so there is no infrastructure to stand up and nothing to maintain. You pay in credits for the calls you make, one currency across all five endpoints. Your research stays private, your queries are ephemeral, and nothing you track trains a model.

Managed for you

One credit currency across five endpoints
Usage-based credits, pay for the calls you make
Ephemeral data, no model training
Mozilla-documented data practices
robots.txt compliance by default

See it live

Drop in your website and a competitor's, get a brief on how to win.

A full read on any competitor: pricing, positioning, product, hiring, and what their users say, plus how to position against them. Cited to the source.

Built on Tabstack

Rival is an open-source competitive intelligence tool built on Tabstack.

Rival tracks competitors across pricing, docs, changelogs, and social signals on a schedule, and it runs on Tabstack. It is open source and built to be read, endpoint by endpoint, so you can see exactly how each call maps to a competitive intelligence job before you build your own.

PricingChangelogsDocsBlogReviewsGitHub activityCareersSocial signalsTech stackPositioning

See Rival on GitHub →View live demo →

Who builds on this

Teams that track competitors without a CI vendor.

Founders

Current intel on the competition, no analyst required.

Developers

Building competitive intelligence into a product.

Growth & marketing teams

Tracking competitor pricing and positioning.

Agents

A competitive read inside a workflow.

Mozilla-backed

Privacy, Transparency, and Control

Every call runs on a Mozilla-backed platform. The pages you extract, the answers you research, and the tasks you automate stay yours, handled responsibly and never used to train models.

Private by default.Requests and retrieved pages are used to complete your call and support you, then purged. Never sold, never used to train models.

Transparent by design.Mozilla-documented data practices and robots.txt compliance by default. See exactly how every endpoint sources and handles your data.

Yours to control.You set the scope and depth of every call. No retained corpus, no lock-in, just clean data you own.

See exactly how we source and handle data in the documentation.

Mozilla Manifesto