Nanonets Document Intelligence
nanonets.com
- Free plan
- Web
- API
- Self-hosted
What Nanonets Document Intelligence is
Nanonets Document Intelligence reads business documents and returns structured data for downstream agent-driven workflows. It handles PDFs, scans, photos, Word and Excel files, and email attachments without per-vendor templates. The product can classify document types and split files that contain multiple documents before extracting fields, tables, and line items. Field-level confidence scores help route uncertain values to human review. It supports 100+ languages, including mixed-script and handwritten documents, and can output JSON, Markdown, or CSV. Nanonets lists 25+ native connectors, including Gmail, Outlook, Google Drive, Dropbox, SAP, Oracle, NetSuite, Salesforce, Slack, and Jira; REST and webhook support are also available. Deployment is offered via API or in a customer’s own VPC. The product is aimed at document-heavy work such as accounts payable, order management, logistics, and healthcare revenue cycle. A free Starter plan includes $50 in credits, then costs $100/month for 100 credits. Usage pricing is based on workflow block runs, with rates of $0.02 per simple run, $0.10 per standard AI run, and $0.30 per complex AI run.
Who it suits
Nanonets is aimed at teams handling document-heavy processes such as accounts payable, logistics, order management, or healthcare revenue cycle. It offers API access and VPC deployment options for organizations with different deployment needs.
Good to know, and what to watch
- Accepts varied document formats without vendor templates
- Classifies and splits multi-document files
- Routes low-confidence fields for human review
- Supports 100+ languages and three output formats
- Offers API and customer-VPC deployment
- Starter costs $100/month after its initial $50 in credits
- Usage is billed by workflow block run
- Growth and Enterprise pricing requires a quote
Plans
Every plan and priceEnterprise
Growth
Starter
Free plan · paid from $100/mo
The verdict
Nanonets combines document classification, extraction, confidence scoring, and a broad set of connectors. Teams should account for both the Starter plan’s included credits and usage-based run rates.
What the maker publishes
- API workflow automation
- API supports extraction, workflow approval/rejection and review assignment. source
- Business process focus
- Purpose-built for document-heavy business processes including accounts payable, order management, logistics and healthcare revenue cycle. source
- Classification
- Automatically classifies document types and splits multi-document files. source
- Classification and splitting
- The product can classify document types and automatically split multi-document files before extraction. source
- Deployment
- The product page says Document Intelligence is available via API or for deployment in a customer's own VPC. source
- Document processing
- Document Intelligence reads documents such as invoices, purchase orders, claims, and contracts and returns structured, agent-ready data. source
- Document understanding
- Processes invoices, POs, claims, contracts and forms, returning structured data. source
- Extraction and review
- It extracts fields, tables, and line items, provides field-level confidence scores, and routes low-confidence values for human review. source
- Headquarters
- San Francisco, California, USA source
- Human review
- Low-confidence fields can be routed to human review. source
- Input formats
- It accepts PDFs, scans, photos, Word and Excel files, and email attachments without per-vendor templates. source
- Integrations
- Nanonets lists 25+ native connectors, including Gmail, Outlook, Google Drive, Dropbox, SAP, Oracle, NetSuite, Salesforce, Slack, and Jira, plus REST and webhook support. source
- Integrations
- Connects to email, cloud storage, APIs and 25+ business systems. source
- Intended users
- Nanonets describes the platform as purpose-built for document-heavy workflows such as accounts payable, order management, logistics, and healthcare revenue cycle. source
- Languages and output
- The product supports 100+ languages, including mixed-script and handwritten documents, and can output JSON, Markdown, or CSV. source
- On-premises deployment
- Customers can run the deep-learning model on their own servers using Docker. source
- Scale
- Nanonets reports 10,000+ customers and over 1 billion documents processed. source
- Security and compliance
- The Enterprise plan lists HIPAA and SOC 2 compliance, while the product page footer also displays SOC 2, HIPAA, GDPR, and ISO. source
- Security controls
- Enterprise controls include SAML SSO, RBAC, audit logs and private deployment. source
- Support
- Starter includes community support, while Enterprise includes dedicated support and SLAs. source
- Usage pricing
- Pricing is based on workflow block runs, with listed rates of $0.02 per simple run, $0.10 per standard AI run, and $0.30 per complex AI run. source
- Usage pricing
- Billing is per workflow block run, with simple, standard AI and complex AI rates. source
Checked against the maker's pages on 15 Nov 2026.