DISCO eDiscovery Pricing: Ingestion Costs, Hosting Fees, and Review Economics

According to the ComplexDiscovery Winter 2026 eDiscovery Pricing Survey co-produced with EDRM, more than half of industry respondents (54.7%) report basic data hosting rates below $10 per GB per month, while another 30.2% fall between $10 and $20 per GB. Yet when litigation teams examine their actual monthly invoices from enterprise eDiscovery platforms like DISCO (CS Disco), the effective cost per gigabyte often diverges significantly from baseline storage estimates. Understanding how these charges accumulate requires examining the operational components that make up cloud litigation pricing.

DISCO eDiscovery pricing is based on a consumption-driven cloud model combining monthly gigabyte hosting fees, data ingestion rates, and optional per-seat or per-query AI review charges. Rather than purchasing perpetual software licenses or installing local server infrastructure, law firms and corporate legal departments pay for the platform resources their matters consume.

A typical enterprise eDiscovery invoice breaks down into five primary billing categories:

• Data Ingestion and Processing: The cost to ingest raw custodian collections, extract container files, de-duplicate records, filter system files, and generate searchable text.
• Monthly Data Hosting: The recurring charge assessed per gigabyte for data maintained inside an active or archived matter database.
• User Licensing and Administrative Seats: Access fees for reviewers, paralegals, litigation support specialists, and outside counsel.
• AI Analysis and Search Intelligence: Advanced machine learning, conceptual search, automated document classification, and generative AI features like DISCO Cecilia.
• Production Deliveries and Exports: The computational and administrative cost of preparing formal court disclosures, applying Bates stamps, burning in redactions, and generating standard load files.

A central element of DISCO's billing framework is the high-water mark model. Actual usage is determined by the greatest amount of data hosted within a customer's database at any point during a calendar month. If a litigation team uploads large custodian collections mid-month and deletes unwanted folders before the billing cycle closes, the monthly invoice reflects the peak volume reached during that cycle rather than the end-of-month baseline. Furthermore, DISCO generally bills customers in advance based on the data quantity and review state at the beginning of the month, applying a true-up adjustment on the subsequent invoice whenever actual peak usage exceeds the initial baseline. Practice managers who understand this structure can avoid costly timing errors during active discovery.

A common point of confusion in eDiscovery billing is the difference between pre-processed collected data and processed reviewable data. When clients collect files from corporate custodians, the initial payload consists of compressed archives, Microsoft Outlook PST files, MBOX containers, mobile device extractions, and loose business documents. This collected volume represents the pre-processed data size.

Before attorneys can review these records, the eDiscovery platform must unpack and index them. The processing engine performs several intensive computational steps:

• De-duplication and DeNISTing: Removing duplicate files across custodians and filtering out standard operating system files using the National Software Reference Library (NSRL) NIST hash database.
• Container Extraction: Extracting nested email attachments, embedded graphics, and compressed zip archives.
• Optical Character Recognition (OCR): Converting scanned TIFF images, non-searchable PDF filings, and handwritten notes into machine-readable text.
• Database and Index Generation: Creating full-text tokens, conceptual search vectors, and metadata mapping tables for rapid querying.

This transformation creates the processed data expansion multiplier. While de-duplication reduces file counts, the extraction of attachments, generation of searchable text, and creation of reviewable images often expands the overall data footprint. In complex litigation, a collection of raw compressed email archives frequently expands to double or triple its original footprint once full-text search indexes, extracted attachments, and viewer files are generated.

The financial consequence of this expansion is substantial. When vendors bill monthly hosting fees against processed data rather than collected data, a firm paying a baseline hosting fee is effectively paying double or triple that rate on a monthly basis for every gigabyte originally collected from the client.

In February 2026, DISCO announced an all-inclusive pricing framework that eliminates separate ingestion and processing charges in favor of a single per-gigabyte price assessed on processed data. While eliminating separate processing line items simplifies cost tracking, the underlying economic reality remains unchanged: because the monthly recurring charge scales directly with processed gigabytes, litigation teams that fail to cull raw data prior to upload pay recurring hosting premiums on non-responsive system files and redundant email chains.

Managing litigation data requires balancing active review accessibility against long-term storage costs. Most eDiscovery matters span months or years, yet only a small fraction of ingested documents are reviewed during depositions or trial prep. Enterprise platforms therefore offer distinct storage tiers to accommodate different phases of the litigation lifecycle.

Modern litigation platforms typically define three primary storage states:

• Active Review Storage: High-performance databases where documents are immediately available for concurrent attorney review, native viewing, real-time tagging, redaction, and search queries. This tier carries the highest monthly per-gigabyte rate.
• Early Case Assessment (ECA) and Nearline Storage: Secondary storage tiers designed for preliminary culling, search term evaluation, and custodian analysis. Documents in ECA can be searched and filtered, but full native viewing, batch tagging, and production tools are restricted until documents are formally promoted to active review.
• Cold Storage and Vaulting: Long-term preservation storage for closed, stayed, or post-trial matters. The database index and document pointers are preserved in an archived state at a fraction of active review costs, but reviving the matter requires an unvaulting delay.

The primary financial risk in multi-tier storage is what practice managers call the highest rate trap. In DISCO's billing framework, data located in more than one review state at any point during a calendar month is generally charged at the highest applicable rate for the entirety of that month.

Consider a practical scenario: a litigation support team ingests a broad custodian collection into an Early Case Assessment tier at the beginning of the month. Later in the month, the case team identifies a subset of responsive documents and promotes them into the Active Review database for attorney tagging. If the promotion process or matter architecture treats the entire matter volume under the active review classification for that billing cycle, the firm incurs full active hosting fees across the entire hosted footprint for the month.

To protect litigation budgets from unnecessary storage inflation, legal operations managers should implement three operational controls:

1. Schedule Promotions at Billing Boundaries: Whenever possible, execute bulk promotions from ECA to active review on the first business day of a billing cycle rather than mid-month.
2. Isolate Custodian Sets in Dedicated Workspaces: Use separate matter databases for broad custodian collections and narrow active review sets to prevent unpromoted data from triggering active review rates.
3. Establish Prompt Post-Settlement Vaulting: Implement a mandatory closeout procedure that shifts settled or inactive matters into vault storage within thirty days of case resolution.

Document review by legal personnel remains the single largest cost driver in modern litigation, routinely consuming the vast majority of total discovery expenditures. Contract attorneys and junior associates typically review between 40 and 60 documents per hour. With contract reviewers and firm associates examining documents file by file, manual review across hundreds of thousands of records rapidly generates substantial billing overhead that often rivals the amount in controversy.

To curb these costs, legal technology has evolved through several generations of automated review:

• Predictive Coding and TAR 1.0: Early Technology-Assisted Review relied on sample sets and expert training runs to predict relevance across document populations.
• Continuous Active Learning (CAL or TAR 2.0): Machine learning algorithms continually update relevance rankings as human reviewers code documents, pushing high-probability responsive files to the top of review queues.
• Generative AI and Agentic Discovery: Modern tools like DISCO Cecilia introduce large language model agents capable of answering complex factual questions, summarizing lengthy deposition transcripts, generating case chronologies, and tagging documents based on natural language instructions.

The economics of generative AI in eDiscovery depend heavily on vendor packaging. Across the legal tech market, pricing models for generative AI review remain diverse. Many vendors charge per-token, per-query, or variable per-document surcharges for generative AI features. For a large document collection, these volume-based AI surcharges can add substantial, unpredictable overhead to an invoice.

DISCO differentiates its offering by bundling its Cecilia AI capabilities, including natural language Q&A, deposition summarization, and timeline construction, directly into its all-inclusive per-gigabyte platform fee. For litigation teams handling document-heavy commercial disputes or complex investigations, this bundled structure offers two clear financial advantages:

• Uncapped Query Exploration: Attorneys and paralegals can query case facts, test legal theories, and generate deposition summaries repeatedly without worrying about metering overages or token consumption.
• Elimination of Per-Document Review Surcharges: The firm avoids secondary review fees when using AI to prioritize custodian records or summarize testimony.

However, legal operations teams must evaluate the reciprocal trade-off. An all-inclusive model incorporates the cost of sophisticated AI infrastructure into the baseline per-gigabyte hosting rate. For routine litigation matters where generative AI queries are unnecessary, a firm with a high baseline hosting fee may end up paying for advanced AI capabilities that its case teams rarely touch.

A primary reason law firms and corporate legal departments face inflated eDiscovery bills is that raw custodian collections are dumped into enterprise discovery platforms prematurely. When paralegals upload unvetted email archives, entire custodian hard drives, and redundant corporate shares directly into an eDiscovery tool, the firm immediately triggers processing multipliers and recurring per-gigabyte hosting fees on files that should have been excluded during initial collection.

Before committing data to specialized platforms like DISCO or RelativityOne, litigation teams need a secure, organized environment for evidence intake, client collection, and preliminary custodian triage. Storing raw evidence on local office network drives or external hard drives creates compliance vulnerabilities and lacks defensible tracking. Conversely, commodity consumer cloud storage tools like Google Drive or Dropbox lack matter-centric isolation, granular legal access permissions, and structured document extraction tools.

Fast.io for legal teams provides a secure cloud workspace platform designed to sit at the front end of litigation workflows, enabling teams to stage documents, manage client intake, and perform early data triage before incurring enterprise processing costs.

Legal teams use Fast.io to organize files into client-centric and matter-centric workspaces:

• Matter-Centric Organization: Firms can create dedicated matter workspaces for each client engagement or litigation matter. Custodian collections, witness disclosures, subpoena responses, and financial records stay separated and organized under granular workspace boundaries.
• Branded Receive Shares for Client Collection: Instead of chasing corporate clients for email attachments or distributing physical thumb drives, litigation teams send branded client portals via Receive share links. Clients drag and drop bank records, contracts, and email archives directly into the designated workspace without creating an account or navigating complicated login walls. Share links can be configured with durable access or specific expiration dates to close collection windows defensively.
• Audit Logging and Chain of Custody: To maintain defensibility during initial collection, Fast.io records every upload, move, download, and file access event in an append-only audit log. Granular permissions at the organization, workspace, folder, and file levels ensure that sensitive financial disclosures and witness materials are accessible only to assigned counsel.
• Structured Data Extraction with Metadata Views: To evaluate document collections before paying for eDiscovery processing, litigation teams use Metadata Views. Attorneys and paralegals describe the fields they want extracted in natural language, such as Agreement Date, Counterparty, Total Claim Amount, Governing Jurisdiction, or Invoice Total. Fast.io automatically extracts these structured fields from PDFs, scanned documents, Word files, and spreadsheets into a filterable, sortable view. This structured extraction allows teams to identify relevant documents and cull irrelevant file sets before uploading them into an eDiscovery platform.
• Collaborative Notes for Custodian Tracking: Real-time Collaborative Notes allow litigation teams to coordinate custodian interview notes, collection logs, and discovery tracking checklists in one shared space alongside matter files.

Fast.io runs on cloud infrastructure partners, including Google Cloud Platform and Cloudflare, that are certified to industry-leading security standards. File data is protected with end-to-end encryption in transit and at rest.

Every organization starts with a 14-day free trial, which requires a credit card. | Plans are Starter at $29/mo, Business at $99/mo, and Growth at $299/mo on Fast.io pricing.

Stage and structure litigation documents before costly eDiscovery

Collect custodian files with secure branded portals, extract structured metadata automatically, and organize matter records in dedicated cloud workspaces. Start your 14-day free trial.

Start 14-Day Trial

Litigation technology leaders evaluating cloud eDiscovery solutions frequently compare DISCO against its two primary cloud competitors: RelativityOne and Everlaw. While all three platforms deliver sophisticated document review and processing capabilities, their underlying commercial architectures reflect contrasting approaches to cost predictability, user licensing, and AI deployment.

Understanding the structural differences between these platforms helps firms align software contracts with their specific case profiles:

DISCO (CS Disco)

DISCO centers its commercial model on transparent, all-inclusive pricing based on processed data volume.

• Licensing Structure: Ingestion fees, user seat licensing, and Cecilia AI features are bundled into a single per-gigabyte price on processed data.
• Billing Mechanics: Monthly bills are calculated based on the high-water mark of hosted data across active review and ECA/vault tiers, with advance billing and true-up reconciliations.
• Key Advantages: High predictability on user counts; unlimited reviewers and AI interrogations without line-item metering surcharges.
• Key Drawbacks: Processed data expansion can inflate hosting costs; the highest applicable rate rule requires careful management when moving data between review states.

RelativityOne

RelativityOne is the cloud-native iteration of the legal industry's longstanding standard platform, offering extensive customization through its partner ecosystem.

• Licensing Structure: RelativityOne typically employs a capacity-based licensing model that combines user licensing, active review storage, and repository (cold) storage allocations. Flex models allow firms to scale capacity as matters require.
• Billing Mechanics: Data is clearly bifurcated between Active Review and Repository Workspace storage, with significant price reductions for data stored in repository mode. Processing is metered based on volume or included in tiered annual subscriptions.
• Key Advantages: Deep customization, massive third-party application ecosystem, and highly competitive cold storage rates for long-term multi-terabyte archiving.
• Key Drawbacks: Invoice complexity can be high, with separate line items for user subscriptions, active storage, repository storage, and advanced generative AI applications (such as Relativity aiR).

Everlaw

Everlaw is a modern, browser-first litigation platform known for high-speed document processing, intuitive user experience, and integrated storytelling tools.

• Licensing Structure: Everlaw operates on a tiered per-gigabyte monthly hosting model that includes processing, document production, predictive coding, and unlimited user seats.
• Billing Mechanics: Per-gigabyte hosting rates scale down automatically as a firm's total hosted volume increases across its matters.
• Key Advantages: Clean pricing transparency with zero user seat fees, no separate processing fees, and integrated deposition and transcript management tools.
• Key Drawbacks: While traditional predictive coding is included, advanced generative AI review features may require distinct credit allocations or usage agreements depending on contract terms.

Contract Negotiation Strategies for Legal Operations

When negotiating eDiscovery contracts with DISCO or its competitors, legal operations managers should focus on four contractual terms:

1. Define the Measurement Standard: Confirm in writing whether hosting charges apply to collected pre-processed data or expanded post-processing volume.
2. Negotiate Realistic Tier Commitments: Avoid overcommitting to annual volume tiers. Aim for contract minimums that cover your core predictable baseline, with pre-negotiated burst rates for unexpected high-volume litigation.
3. Establish Immediate Tier Transitions: Require that data moved from active review to vaulting or cold storage transitions immediately upon status change, eliminating contractual clauses that charge active rates for the full billing cycle.
4. Pre-Stage Collections Outside the Platform: Implement an initial staging protocol using a matter workspace platform like Fast.io to cull system files, extract structured metadata, and verify responsive custodians before data enters the eDiscovery processing pipeline.