Pricing

AI-ready data pricing for SourceFusion processing and curated datasets.

DeepWave SourceFusion turns documents, web pages, public feeds, APIs, tables, geospatial/time-series data, and unconventional sources into RAG-ready chunks, model features, clean records, API payloads, lineage, and quality fields. Document-style inputs use page pricing. Curated datasets, APIs, and feeds use product access or private deployment terms.

Plans

Start with page processing. Expand into datasets, APIs, and private pipelines.

A page is the pricing unit for PDFs, reports, scanned files, HTML, and document-style web sources. Structured datasets remain priced as prepared products or enterprise feeds.

Free

SourceFusion Free

Start processing documents and web sources into AI-ready outputs before committing to a paid plan.

$0
15,000 pages included
  • 15,000 included pages with no paid commitment
  • PDF, HTML, report, OCR, and file processing
  • RAG chunks, metadata, citations, and API-ready JSON examples
  • Schema, quality, lineage, and limitation fields
Open free processing
Dataset packs

Prebuilt Dataset Access

Provision prepared DeepWave datasets such as Extreme Weather, Antimicrobial Resistance (AMR), and Contrail Research with export files and manifest records.

From $397
Per prepared dataset package
  • CSV, JSON, JSONL, API payload, and model-feature exports
  • Source lineage and dataset dictionary
  • Quality fields and validation notes
  • Account-based entitlement after payment confirmation
Choose dataset
Business

Business and Private Deployment

Dedicated SourceFusion pipelines, private deployment, custom geography, API/feed delivery, marketplace procurement, or customer-specific source domains.

Custom
Scoped by source mix, volume, cadence, and deployment boundary
  • Dedicated source acquisition and harmonization
  • Customer VPC, private endpoint, or private cloud offer
  • Custom schemas, embeddings, and model features
  • API/feed SLAs, security, deployment, and renewal terms
Scope custom pipeline
All plans include

AI-ready outputs, not raw downloads.

Every DeepWave output is designed for downstream AI systems: page-level extraction when the input is document-like, normalized records when the input is structured, and retrieval-ready chunks, features, provenance, and quality records across both.

IngestDocuments, web pages, public sources, APIs, tables, and approved unconventional sources are collected and timestamped.
NormalizeDifferent cadences, location grains, and source formats are harmonized into consistent rows and geographies.
StructureDatasets ship with schemas, dictionaries, quality scores, limitation fields, and release metadata.
ChunkRAG-ready JSONL records include stable IDs, concise text, metadata, and source lineage for retrieval systems.
FeatureFeature tables expose numeric, categorical, temporal, and geospatial data for AI workflows.
PriceDocument-style processing is $0.03/page after 15,000 included pages. Dataset packs, APIs, and feeds use their listed product terms.
SecureAccount entitlement, customer-controlled credentials, private deployment options, and marketplace procurement.
ValidateData-quality fields, lineage, generated timestamps, and package manifests support review before automation.
DeliverCSV, JSON, JSONL, API payloads, AWS Marketplace, AWS Data Exchange, Azure Marketplace, and private delivery paths.
Pricing basis

Use the pricing model that matches the product.

A page is meaningful when SourceFusion parses a document. A record is meaningful when DeepWave maintains an AI-ready data product.

Document pagesPDF pages, scanned pages, report pages, HTML pages, images requiring OCR, and long document sections processed as pages.
RecordsSource-linked observations with time, geography, entity, source family, measurement, quality, and limitation fields.
RAG chunksRetrieval-ready JSONL entries with text, metadata, citation fields, source lineage, and stable identifiers.
Feature rowsAI-ready rows for forecasting, scoring, alerting, enrichment, and feature-store workflows.
FeedsScheduled updates priced by geography, refresh cadence, source volume, delivery target, and support level.
Dataset packsPrebuilt packages such as Extreme Weather, Antimicrobial Resistance (AMR), and Contrail Research with fixed catalog access and optional private delivery.
Prepared datasets

Current catalog pricing uses the live product configuration.

These are the standard packages available for preview and account-based access. Private marketplace terms can wrap the same products for enterprise procurement.

Dataset Use AI-ready outputs Price
Antimicrobial Resistance (AMR) Public Source Signal Dataset
Public Antimicrobial Resistance (AMR) evidence, markers, source lineage, and outbreak-watch scoring
Antimicrobial Resistance (AMR) outbreak-watch signal built from public resistance evidence and source coverage CSV, JSON, JSONL, model features, manifest, lineage $397 Preview
Extreme Weather Signal Dataset
Location-window prediction features for DC Metro and selected metros
Localized extreme-weather risk can be compared by grid cell and time window CSV, JSON, JSONL, model features, manifest, lineage $397 Preview
PermitPulse Construction Activity Dataset
Permit, project, license, and public-space activity evidence
Construction activity is measurable here CSV, JSON, JSONL, model features, manifest, lineage $397 Preview
GrantSignal Funding Opportunities Dataset
Funding opportunities, awards, notices, and pursuit evidence
Funding signals can be compared here CSV, JSON, JSONL, model features, manifest, lineage $397 Preview
Antimicrobial Resistance (AMR) Priority Pathogens Dataset
Observation-first antimicrobial resistance dataset
Measured antimicrobial resistance evidence is available here CSV, JSON, JSONL, model features, manifest, lineage $497 Preview
Contrail Research Dataset
Aviation-weather evidence for contrail research
Flight and atmospheric conditions can be compared here CSV, JSON, JSONL, model features, manifest, lineage $697 Preview
Cloud delivery

Procure through AWS or Azure when the buying path requires it.

Marketplace packages use the same dataset products, AI-ready outputs, lineage, and entitlement model with private-offer commercial terms.