Unstructured Technologies
Transform unstructured data into AI-ready formats efficiently.. [Contact for Pricing]
Last verified:
Visit Unstructured Technologies
What is Unstructured Technologies?
Unstructured Technologies is an ETL+ platform designed specifically for GenAI data that transforms complex, unstructured data into clean, structured data ready for AI applications. The platform processes 64+ file types including PDFs, Word documents, PowerPoint presentations, images, audio, and video files, routing data through dynamic transformation and enrichment pipelines to deliver high-quality output to LLMs.
Key features include partitioning, chunking, enrichment, and embedding capabilities with multiple strategies like character-based, page-based, similarity-based, and contextual chunking. The platform offers 40+ connectors to sources like AWS S3, Google Drive, SharePoint, Slack, and databases, plus 20+ destination connectors to vector stores like Pinecone, Weaviate, and Chroma. Users can access the platform through a no-code UI for teams who prefer visual interfaces or via API for engineers needing programmatic control.
Unstructured Technologies is built for enterprises and development teams building GenAI applications, RAG systems, and LLM-powered solutions. It is trusted by 87% of the Fortune 1000 and over 60,000 organizations globally. The platform serves industries including banking & finance, healthcare & life sciences, legal tech & compliance, CPG, and automotive/airline sectors.
The platform includes built-in security and compliance (SOC 2 Type 2, HIPAA, GDPR, ISO 27001, FedRAMP, CMMC 2.0 Level 2), role-based access control, 24/7 connector maintenance, and supports integration with major AI providers like OpenAI, Anthropic, Azure AI Studio, Amazon Bedrock, Google Gemini, IBM watsonx, and NVIDIA.
Unstructured Technologies pricing
Pricing model: Freemium
Unstructured offers three pricing tiers: Let's Go (Free): 15,000 free pages with no expiration date, no minimums, completely free, includes all features, single user with single workspace hosted on Unstructured's cloud. Pay-As-You-Go: $0.03 per page, flat rate for any file type and pipeline, pay only for what you process, no minimums, no maximums, no commitment, no hidden fees, includes all features. Business (Custom Pricing): Built for teams needing privacy, control, and security. Includes multi-user accounts, all features, and deployment options: Business SaaS (cloud-hosted), Dedicated Instance (isolated VPC on Unstructured's cloud), or In-VPC (your own VPC on your cloud). Custom pricing tailored to needs. Page calculation: PDF/PPTX/TIFF = 1 page per page/slide/image; DOCX with metadata uses metadata; other files = size/100KB; non-file data = 100KB chunks.
Unstructured Technologies pros
- Processes 64+ file types including PDFs, Word docs, PPTs, images, audio, and video
- 15,000 free pages with no expiration on Let's Go plan
- No-code UI accessible for teams without heavy coding experience
- Full API access for engineers needing programmatic control
- 40+ source connectors including S3, Google Drive, SharePoint, Slack
- 20+ destination connectors to vector stores like Pinecone, Weaviate, Chroma
- Multiple chunking strategies: character, page, similarity, and contextual
- Built-in SOC 2 Type 2, HIPAA, GDPR, ISO 27001 compliance
- 24/7 connector maintenance as systems evolve
- Access to vision transformer models and fine-tuned OCR models
- Generative OCR enrichment and image/table description enrichment
- Role-based access control for enterprise security
- Deployment options: SaaS, Dedicated Instance, or In-VPC
- Zero data retention policy for enhanced security
- Trusted by 87% of Fortune 1000 and 60,000+ organizations globally
- Incremental processing and change detection intelligence
- Support for custom models for document hierarchy and element classification
- Full ETL orchestration with smart document routing
- Integrates with OpenAI, Anthropic, Azure AI Studio, Bedrock, Gemini, NVIDIA
Unstructured Technologies cons
- $0.03 per page on Pay-As-You-Go can become expensive at scale
- Business plan pricing is custom and requires contacting sales
- Some advanced features like Custom Enrichments only available in VPC
- Video-to-Text and Speech-to-Text only available in In-VPC deployment
- Structured Data Extraction feature still coming soon
- Custom Plugins for Specialized Pipelines only available In-VPC
- Custom Model Hosting for LLMs/VLMs only available In-VPC
- Some preview connectors require enabled-on-request access
- Page calculation for non-standard file types uses 100KB division
Frequently asked questions about Unstructured Technologies
What is Unstructured and what does it do?
Unstructured is an ETL+ platform for GenAI data that automatically transforms complex, unstructured data into clean, structured data for GenAI applications. It parses, chunks, enriches, and embeds data from 64+ file types, routing data through dynamic transformation pipelines to deliver the highest quality output to your LLM continuously, effortlessly, and automatically.
What file types does Unstructured support?
Unstructured supports over 64 file types including 50+ document and image formats (PDF, DOCX, PPTX, HTML, JPEG, PNG, TIFF, CSV, XML) and 20+ audio/video formats (MP3, MP4, WAV, AVI, FLAC, MOV, WebM). It also processes non-file data with pages calculated as 100KB chunks.
What is the free tier and how does it work?
The Let's Go plan offers 15,000 free pages with no expiration date, no minimums, and is completely free. It includes full access to every feature in the platform, single user access, and single workspace hosted on Unstructured's cloud infrastructure.
How is pricing calculated for Pay-As-You-Go?
Pay-As-You-Go costs $0.03 per page at a flat rate for any file type and pipeline. Pages are calculated as: PDF/PPTX/TIFF = 1 page per page/slide/image; DOCX with page metadata uses that metadata; other files = file size divided by 100KB; non-file data = 100KB of incoming data to be processed.
What deployment options are available?
Unstructured offers SaaS Cloud-Hosted (Let's Go, Pay-As-You-Go, Business SaaS), Dedicated Instance (hosted in isolated VPC on Unstructured's cloud for additional security), and In-VPC (hosted within your own VPC on your cloud infrastructure). Dedicated Instance and In-VPC are Business Plan only.
What security and compliance standards does Unstructured meet?
Unstructured adheres to SOC 2 Type 1, SOC 2 Type 2, HIPAA, GDPR, ISO 27001, FedRAMP, and CMMC 2.0 Level 2 compliance standards. Features include data encrypted in transit, permission-based access with role-based access control (RBAC), secure connector authentication credential handling, zero data retention, and data only processed within your company's dedicated infrastructure for Business plans.
What chunking strategies does Unstructured offer?
Unstructured offers five chunking strategies: Chunk by Character, Chunk by Title, Chunk by Page, Chunk by Similarity, and Contextual Chunking. The API provides access to by-page and by-similarity chunking strategies as a benefit over open source.
What connectors and integrations are available?
Unstructured has 40+ connectors including 20+ sources (AWS S3, Google Drive, SharePoint, Slack, Salesforce, Azure Blob Storage, Box, Dropbox, Notion, Confluence, Jira, Discord, GitHub, GitLab, MongoDB, PostgreSQL, Snowflake) and 20+ destinations (Pinecone, Weaviate, Chroma, Neo4j, Elasticsearch, Qdrant, Redis, MongoDB, AWS Delta Table, Azure AI Search). It integrates with OpenAI, Anthropic, Azure AI Studio, Amazon Bedrock, Google Gemini, IBM watsonx, NVIDIA, and Together.ai.
What enrichment features does Unstructured provide?
Enrichment features include Metadata Extraction, Named Entity Recognition (NER), Generative OCR, Image Description, Table Description, and Table to HTML conversion. The platform also offers custom models for document hierarchy and element classification, third-party model integration, and custom enrichments (VPC only).
How do I get support or contact Unstructured?
For technical support on Unstructured accounts, email [email protected]. For sales inquiries or to upgrade from Let's Go/Pay-As-You-Go to Business, email [email protected]. The platform also offers Slack Community support, and Business plans include a Personal Support Representative and email support.