RESEARCH API V4 IS LIVE


Give your LLMs real-time access to the entire web. Structured JSON output in under 30 seconds.

Get API Key
example.tsresponse.json
// 1. Define your schema
const schema = {
company_name: "string",
revenue_growth: "string",
key_risks: ["string"]
}
// 2. Call UnforgeAPI
const res = await fetch('https://www.unforgeapi.com/api/v1/deep-research', {
method: 'POST',
headers: { 'Authorization': 'Bearer uf_your_api_key' },
body: JSON.stringify({
query: "Microsoft Q3 2026 Results",
mode: 'schema',
preset: 'stocks',
schema
})
})
200 OK
// Response (~30s)
{
"company_name": "Microsoft Corporation",
"revenue_growth": "+17% YoY",
"key_risks": [
"AI regulation (EU AI Act)",
"Data center capacity constraints"
],
"sources": [
{ "title": "Microsoft Q3 2026", "url": "..." }
],
"meta": { "latency_ms": 28450, "preset": "stocks" }
}
~30s
Deep Research
100%
Structured JSON
6
Domain Presets
V4 Pipeline Engine

How Deep Research
Works

A simplified multi-stage AI pipeline. From raw web search to structured data synthesis in three powerful steps.

1. Search

Concurrent fetching of authoritative web sources, gathering raw content across trusted domains instantly.

2. Verify & Check

Iterative fact-checking agent detects outdated data and filters noise to refine search queries for precision.

JSON
Done
Verified

3. Reason & Render

Structuring data into a readable format immediately with citations and source links ready for ingestion.

The Missing Infrastructure for
Autonomous Agents

Why raw scraping fails and structured knowledge succeeds. A verifiable engine for the next generation of AI.

Unreliable Inputs

Flaky Scrapers

HTML parsing errors

Broken RAG

Hallucination risk

API_ACTIVE

Knowledge Layer

v2.0.4 stable

Reliable Outputs

Structured JSON

Schema-compliant data

Verified Facts

With citation links

Ready to build reliable agents?

Stop debugging scrapers and start building intelligence. Get your API key today.

Next Gen Infrastructure

Everything your agent needs tounderstand the world.

Power your AI with structured data, real-time extraction, and seamless integrations. A unified system designed for the next generation of intelligent agents.

Custom Schemas

Stop parsing markdown manually. Define the exact JSON structure you need, and we'll fill it with clean, real-time data from any source.

Read documentation
schema.json
// Define your JSON schema
{
"query": "Tesla stock analysis",
"mode": "schema",
"preset": "stocks",
"schema": {
"company_name": "string",
"current_price": "number",
"price_change": "string",
"key_risks": ["string"],
"analyst_rating": "string"
}
}
agentic.json
// Enable iterative reasoning
{
"query": "2026 AI market trends",
"mode": "report",
"preset": "tech",
"agentic_loop": true
}
// Iterates up to 3x to verify data
// Eliminates outdated predictions

Agentic Reasoning

Iterative research that validates sources and eliminates hallucinations. Our reasoning layer detects outdated data and refines queries automatically for accurate, current results.

Read documentation

Multi-Query Compare

Compare multiple topics in a single researched response. Aggregate data points effortlessly into a unified view for complex decision making.

POST /api/webhook
200 OK
timestamp:1679823422
event:"job_completed"
payload:{ ... }
Transmission active

Webhook Delivery

Async processing? We'll POST the JSON payload to your endpoint the millisecond it's ready. Reliable, retriable, and built for scale.

PRICING

Simple, Transparent Pricing

Deep research API with 3-iteration agentic loop. All tiers fully managed - we handle the AI provider costs.

Free

$0

Perfect for testing

  • 10 deep research / month
  • Community support
Most Popular

Pro

$29/month

For growing teams

  • 100 deep research / month
  • Priority support

Expert

$200/month

For high-volume production

  • 800 deep research / month
  • Dedicated account manager
  • SLA guarantee
All plans include our universal router and 99.9% uptime SLA.

Frequently Asked Questions

Ready to Stop Overpaying?

Join the developers who cut their AI costs by 70% with intelligent query routing.