Description
Hyperresearch provides a robust AI deep research platform designed to deliver durable, verified reports. It integrates a sophisticated AI search engine that pulls data from over 90 sources, including SEC filings, Census, BLS, and PubMed. A core capability is its rigorous citation verification process, where every claim in a report is matched word-for-word against its original source, or explicitly marked as "not found."
Read More
The platform operates through a multi-stage process, starting with decomposing a user's question into atomic components and partitioning it into several research chapters. It then conducts a wide sweep, prioritizing the user's existing vault before performing planned searches across various lenses. This is followed by a contradiction graph analysis, loci analysis to determine depth, and parallel depth investigations. The process culminates in cross-locus reconciliation, source tension identification, and a corpus critic stage to fetch any missing gaps before writing.
Reports are written in a single pass, drawing from an evidence map of fully read sources. Each report includes a comprehensive plan, and a final claim check stage verifies every cited figure against its source, correcting any discrepancies. Users receive a detailed report, a full list of sources, an extracted claims table, and a verification receipt, ensuring transparency and accuracy.
Hyperresearch offers two main service tiers: Light and Deep. A Light run costs $9 and provides an answer-first report of 2,500 to 4,000 words, typically taking 30 to 40 minutes, with cited claims checked. A Deep run costs $49, involves 150 to 200 sources, takes 3 to 5 hours, and includes a verification receipt detailing citation checks. The first Deep run is free for new users who add a card. The platform also offers an API for integrating its research capabilities into other tools, allowing users to connect their existing AI assistants like Claude, Cursor, or ChatGPT to their Hyperresearch vault.
- CategoryAI Research Assistants
- Pricing ModelPay-as-you-go
- DeploymentCloud
- PlatformsWeb
Hyperresearch's Core Features
Durable, verified AI search reports
Integration with 90+ data sources (SEC filings, Census, PubMed)
Every claim matched word for word against its source
Automatic saving of all sources to a personal vault
Multi-stage research pipeline (Decompose, Chapter Partition, Width Sweep)
Contradiction graph and Loci analysis for deep investigation
One-pass report writing from an evidence map
Comprehensive claim checking and verification receipts
Two report sizes: Light ($9) and Deep ($49)
API access for integration with other AI agents (Claude, Cursor, ChatGPT)
Workspace vault for searchable notes and sources
Citation verification API for external documents
Pay-as-you-go pricing model
Monthly spend cap option
Export vault as Markdown
How to use Hyperresearch?
Create Account: Sign up at console.hyperresearch.ai and add a card to get your first Deep run free.
Frame Question: Submit your research question to initiate the process.
Select Run Size: Choose between a Light run ($9) for focused questions or a Deep run ($49) for broader investigations.
Review Report: Receive a comprehensive report with a source list, claims table, and verification receipt.
Access Vault: Explore your workspace vault containing all sources and interim notes from your runs.
Integrate Agents: Connect your existing AI assistants like Claude or ChatGPT to your vault via the hosted MCP server.
Hyperresearch's Use Cases
- Academic Research
- Market Analysis
- Due Diligence
- Content Creation
- Legal Research
- Competitive Intelligence
- Policy Analysis
- Medical Literature Review
