FutureSearch Agent

Research, then analyze

An agent researches each row on the web, then analyzes what it finds. Add columns of data you don't have yet.

Try Research freeGive to your AI

▶ 2-min demo video coming soon

Diagram showing a company column enriched with annual pricing data: Figma $144, Notion $96, Linear $120, Airtable $60
1–11¢per row
accuracy verifiedin Deep Research Bench
💰

company → annual_price, tier_name

Enrich SaaS Pricing

Research pricing pages for hundreds of products. Extract tier names, annual prices, and feature lists as structured columns.

$6.68 • 99.6% success • 246 products

🏷️

job_title → category, seniority

Classify Job Postings

Add category, seniority level, and confidence columns to job listings using LLM classification.

$1.74 • 100% success • 200 postings

📦

package → days_since_release, contributors

Research Package Metadata

Look up days since last release, contributor counts, and other metrics for PyPI packages from the web.

$3.90 • 1.3¢/row • 300 packages

Give your AI a team of agents

Claude Code

claude mcp add futuresearch --scope project --transport http https://mcp.futuresearch.ai/mcp

Then ask Claude to research your data.

Python SDK

pip install futuresearch

from futuresearch.ops import agent_map
result = await agent_map(
  task="Find the annual price
    of the lowest paid tier",
  input=products_df,
  response_model=PricingInfo
)

Pricing

Start with $20 in free credits. No credit card required. Pay only for what you use—costs scale with research complexity.

TaskRowsCost/rowSuccess
SaaS pricing lookup2462.7¢99.6%
Job classification2000.9¢100%
Package metadata3001.3¢—

Why costs vary

Every row gets its own web research agent. Agents have degrees of freedom. They spend more tokens doing more research for harder tasks. Simple lookups finish quickly; complex research requires multiple page visits and reasoning steps.

Resources

Tutorials

SDK Docs

Case Studies