The Great eDiscovery Price Reset: How Free AI, All-Inclusive Pricing, and Cloud Competition Are Finally Making Justice Affordable

By Sid Newby | April 2026

In twenty-plus years of building and deploying litigation technology, I've watched one pattern repeat with depressing regularity: powerful tools get built, large firms and corporations buy them, and everyone else gets priced out. The eDiscovery industry perfected this model -- charging per gigabyte for processing, per gigabyte for hosting, per document for review, per hour for consulting, and then layering on AI surcharges when the technology that was supposed to reduce costs arrived. For two decades, the pricing complexity itself became a competitive moat, ensuring that only well-funded parties could afford defensible discovery. But something remarkable is happening in 2026. The walls are coming down. Relativity just made its GenAI review tools free. Everlaw followed within weeks. DISCO collapsed its entire platform into a single transparent price. And the Winter 2026 eDiscovery Pricing Survey shows an industry in the early stages of the most significant pricing transformation since cloud hosting replaced on-premises servers. This is the story of how eDiscovery pricing finally broke -- and why it matters far more than most people realize.

The legacy pricing labyrinth: how we got here

To understand why the current pricing transformation is so significant, you need to understand the Rube Goldberg machine that eDiscovery pricing became over the past two decades.

The traditional eDiscovery pricing model evolved organically from the service bureau era of the early 2000s. When litigation teams first began dealing with electronic evidence at scale, they outsourced nearly everything to specialized vendors who charged for each discrete step in the EDRM (Electronic Discovery Reference Model) workflow. The result was a pricing architecture that looked something like this:

Table 1: Traditional eDiscovery pricing model components. Source: ComplexDiscovery Winter 2026 Pricing Survey.[^1]

A single matter with 100GB of data could easily generate five- to six-figure bills spread across a dozen line items, each with its own opaque calculation methodology. For large litigation, the numbers were staggering: the landmark 2012 RAND Corporation study estimated that each gigabyte of ESI cost approximately $18,000 to produce through the full discovery lifecycle.^[2] While technology has driven that number down dramatically, the architectural complexity of the pricing model persisted -- and with it, the barrier to entry for smaller firms and under-resourced litigants.

The per-gigabyte model had a particularly insidious effect: it penalized data volume rather than work performed. A custodian with a large mailbox cost more to discover than one with a small mailbox, regardless of whether the additional data contained anything relevant. This created perverse incentives to under-collect, under-process, and under-review -- exactly the opposite of what defensible discovery requires.

Figure 1: The evolution of eDiscovery costs per gigabyte, from the $18,000/GB RAND estimate in 2012 to the sub-$100/GB all-inclusive models emerging in 2026.

The three shocks that broke the old model

The eDiscovery pricing transformation of 2026 didn't happen gradually. It was triggered by three near-simultaneous market shocks that collectively made the old pricing architecture untenable.

Shock 1: Relativity and Everlaw make GenAI free

In October 2025, at Relativity Fest, Relativity announced that its aiR for Review and aiR for Privilege generative AI solutions would be included in the standard RelativityOne package at no additional charge starting in early 2026.^[3] This was seismic. Relativity's aiR products -- which use OpenAI's GPT models to perform first-pass document review, privilege detection, and case strategy analysis -- had been premium-priced features that only the largest and best-funded review operations could justify.

Within weeks, Everlaw responded by making its single-use EverlawAI Assistant features free as well, including its Writing Assistant, Deposition Analyzer, and Single Document Review Assistant.^[3] Tasks that didn't run in batch across multiple documents would no longer consume AI credits.

The significance of these announcements cannot be overstated. As one industry analysis noted, "These announcements represent perhaps the most significant step forward in [access to justice] that we've seen in a generation."^[3] The credit-based pricing system that had governed AI features created friction and unpredictability. Attorneys hesitated to adopt AI features due to unclear per-task costs, reserved advanced tools for critical tasks only, and feared financial consequences of inefficient queries. By eliminating per-use charges, Relativity and Everlaw removed the single biggest barrier to AI adoption in document review.

Shock 2: DISCO's all-inclusive pricing collapse

While Relativity and Everlaw were unbundling AI charges, DISCO took a more radical approach: collapsing its entire platform into a single per-GB hosting fee that includes eDiscovery, Cecilia AI (its GenAI assistant), deposition management, timelines, and case analytics.^[4] On the DISCO platform, you can now interrogate your evidence, create a timeline, summarize a deposition, and more on your largest datasets at no added cost.

DISCO's Cecilia Auto Review can process approximately 25,000 documents per hour using plain-English prompts crafted in collaboration with eDiscovery experts -- and this capability is included in the base price.^[4] Compare that to traditional human review at 50-80 documents per hour per reviewer, at $1.50-$3.00 per document, and the economic disruption becomes clear.

Shock 3: Cloud commoditization drives hosting costs down

The third shock was quieter but equally important. Cloud hosting costs have been drifting down 2-5% per year as cloud efficiency gains accumulate.^[1] More significantly, the competitive pressure from all-inclusive platforms like DISCO forced traditional per-GB hosting providers to justify their premium pricing for what has become commodity infrastructure. The Winter 2026 eDiscovery Pricing Survey found that 54.7% of respondents now charge below $10/GB per month for basic hosting, with downward pressure accelerating.^[1]

The new pricing landscape: what the data shows

The Winter 2026 eDiscovery Pricing Survey, conducted by ComplexDiscovery with 53 respondents across law firms (43.4%), vendors (24.5%), and corporations (15.1%), provides the clearest picture yet of an industry in transition.^[1]

GenAI review pricing is fragmenting -- in a good way

The survey reveals that GenAI-assisted review pricing has no consensus model yet. Hybrid models and per-document billing each represent approximately 28% of primary pricing structures. The per-document range of $0.11-$0.50 is emerging as the competitive zone for GenAI review -- a dramatic reduction from traditional human review costs of $1.50-$3.00 per document.^[1]

This fragmentation is healthy. It means the market hasn't yet calcified around a single pricing model, and buyers have leverage to negotiate. More importantly, the per-document cost for AI review is converging toward a level where comprehensive review becomes economically feasible for cases that would have been cost-prohibitive under the old model.

Processing costs continue their long decline

For data processing, 39.6% of respondents report ingestion costs of $25-$75 per GB, while 34.0% report costs below $25/GB.^[1] Compare this to the $35-$75/GB processing charges that were standard just two years ago, and the trend is clear.^[9] The combination of faster hardware, more efficient algorithms, and competitive pressure is pushing processing toward commodity pricing.

The rise of alternative pricing models

Perhaps the most telling signal in the survey is the growing proportion of respondents reporting non-traditional billing models: 18.9-22.6% across processing and hosting categories indicated "alternative pricing" rather than per-GB or per-document structures.^[1] These alternatives include flat-rate case pricing, subscription models, and outcome-based billing -- all of which shift risk away from the buyer and toward the provider.

Table 2: eDiscovery pricing model market share evolution, 2024-2026. Source: Author analysis based on ComplexDiscovery surveys.[^1]

What this means for litigation teams

The pricing transformation has concrete, immediate implications for every organization involved in litigation.

For corporate legal departments: leverage the disruption

In-house legal teams have more pricing leverage than at any point in the last decade. The combination of free GenAI tools from Relativity and Everlaw, all-inclusive alternatives from DISCO, and general downward pressure on per-GB costs creates a buyer's market. Specific actions:

• Renegotiate existing contracts with eDiscovery vendors, particularly any that still charge AI surcharges on top of hosting fees
• Benchmark your per-GB costs against the survey data: if you're paying above $10/GB/month for hosting or above $75/GB for processing, you're overpaying
• Evaluate all-inclusive platforms for matters where predictable budgeting matters more than best-of-breed features
• Demand transparent pricing -- if your vendor can't give you a clear, written explanation of every line item on your invoice, that's a red flag

For law firms: adapt or lose clients

The pricing survey data shows that corporate clients are increasingly sophisticated about eDiscovery costs. Firms that continue to pass through opaque vendor charges with markups will lose business to firms that offer transparent, predictable pricing. The firms that thrive will be those that:

• Invest in AI-powered review now that the tools are included at no extra charge -- firms that don't adopt free GenAI capabilities are leaving value on the table
• Offer fixed-fee or capped-fee arrangements for eDiscovery that align with the all-inclusive models clients are seeing from platforms
• Build in-house expertise rather than outsourcing everything to vendors at premium rates
• Use technology to compress review timelines, which reduces hosting costs and accelerates case resolution

For plaintiff firms and small practices: the opportunity of a generation

This is where the pricing transformation matters most -- and where I'm most passionate about the implications.

For twenty years, small plaintiff firms have been systematically disadvantaged in discovery. They couldn't afford the same tools, the same review platforms, or the same AI capabilities as the large defense firms they opposed. A plaintiff's firm handling a complex commercial litigation matter might spend $50,000-$100,000 on eDiscovery before a single deposition was taken, while the defense firm's client could absorb those costs without blinking.

The 2026 pricing reset changes this calculus fundamentally. When GenAI review is free on both Relativity and Everlaw, the technology advantage that large firms enjoyed in document review evaporates.^[3] A solo practitioner with a RelativityOne subscription can now access the same AI-powered review capabilities that AmLaw 100 firms use. DISCO's all-inclusive pricing means a small firm can budget eDiscovery costs with certainty, without fear of spiraling per-document charges.

As the ILS analysis noted, these changes signal recognition that "justice should not be tiered based on who can afford the best technology."^[3] The recommendation for plaintiff firms is clear: adopt and integrate GenAI quickly, train comprehensively on AI prompting and output validation, and partner with plaintiff-focused eDiscovery providers who understand the unique economics of contingency practice.^[3]

The market is growing -- and transforming

The eDiscovery market itself continues its rapid expansion, providing context for why vendors are competing so aggressively on pricing. The global eDiscovery market was valued at approximately $18.73 billion in 2025 and is projected to reach $46.06 billion by 2034, with particularly strong growth in the SME segment.^[5]

This growth is being driven by several converging factors:

• Data volume expansion: Enterprise data continues to grow exponentially, driven by collaboration platforms, cloud storage, and AI-generated content
• Regulatory complexity: New data privacy regulations globally are increasing the frequency and complexity of compliance-related discovery
• AI capability expansion: GenAI is creating new use cases and expanding the scope of what's feasible in discovery, attracting new buyers
• SME adoption: Small and medium enterprises are adopting eDiscovery solutions at a high pace to enhance governance, risk, and compliance programs^[5]

The SME segment is particularly significant. Vendors in the market have been deliberately rolling eDiscovery solutions with the intent to cater to SMEs, recognizing both the market opportunity and the access-to-justice imperative.^[5] This is a fundamental shift from the enterprise-first strategy that dominated the industry for two decades.^[10]

The unresolved questions

The pricing transformation is real, but it's not without complications. Several critical questions remain unresolved.

What happens to review quality when AI is free?

When AI review is included at no charge, the incentive to use it on every matter increases dramatically. That's generally positive, but it also means more AI-assisted review is being done by teams with less experience in validating AI outputs. The Winter 2026 survey notes that exception handling, quality control, and contract structures around GenAI services remain underdeveloped.^[1]

The risk is that free AI creates a false sense of security. A tool that achieves 90%+ recall is still missing documents. The question is whether the cost savings from AI review are being reinvested in human quality control -- or whether firms are pocketing the difference and reducing oversight. The hallucination risks documented in the courts make this question urgent.^[6]

Will pricing transparency survive the AI hype cycle?

The current pricing disruption is partly driven by competitive pressure: Relativity made GenAI free, so Everlaw had to follow. DISCO's all-inclusive model puts pressure on per-GB competitors. But competitive pricing wars eventually end. When they do, will the industry return to opaque, multi-layered pricing? Or will the transparency genie stay out of the bottle?

History suggests caution. The legal technology industry has repeatedly found ways to re-introduce complexity into pricing. AI inference costs are real -- someone is paying for the GPU compute time when aiR processes a million documents. If those costs aren't reflected in per-use charges, they'll be absorbed into base subscription prices, which may rise significantly over time.

Where does forensic collection fit?

While processing and review costs are falling, forensic collection pricing has remained stubbornly high, with 56.6% of respondents reporting rates of $250-$350 per hour and expert witness testimony commanding over $550 per hour in 26.4% of cases.^[1] Collection requires specialized expertise, onsite presence, and chain-of-custody documentation that doesn't easily scale or automate.

This creates an emerging gap: the technology for reviewing evidence is becoming cheap, but the expertise to defensibly collect it remains expensive. For smaller matters and under-resourced parties, collection costs may become the new bottleneck that prevents affordable access to discovery.

How do courts handle AI-reviewed production?

Courts are still developing standards for AI-assisted review. While TAR has been widely accepted since the landmark Da Silva Moore v. Publicis Groupe decision in 2012 and the Rio Tinto approval in 2015, GenAI-assisted review is newer and less tested.^[7] The validation frameworks built into Relativity's aiR -- which mirror the statistical principles established for TAR workflows -- are a good start, but court-by-court acceptance will take time.^[8]

Litigation teams using GenAI review should document their methodology, validation statistics, and human oversight procedures with the same rigor they'd apply to any TAR protocol. The technology may be free, but defensibility isn't optional.

The access to justice imperative

I want to end where I always end -- with access to justice. Because ultimately, eDiscovery pricing isn't just an industry story. It's a justice story.

When discovery costs $18,000 per gigabyte, only the wealthiest litigants can afford comprehensive review. When review costs $3.00 per document, small-value claims become economically irrational to pursue. When AI tools are locked behind enterprise pricing tiers, the technological divide between large and small firms grows wider with each innovation cycle.

The 2026 pricing reset offers the first real hope I've seen in twenty years that this pattern might change. Free GenAI review tools from Relativity and Everlaw don't just save money -- they fundamentally alter the competitive dynamics of litigation. All-inclusive pricing from DISCO doesn't just simplify budgets -- it makes discovery costs predictable enough for contingency-fee practices to manage.

But tools alone aren't enough. The eDiscovery industry -- vendors, law firms, and corporate legal departments alike -- has a responsibility to ensure that pricing transparency becomes permanent, not temporary. The market forces driving costs down today could reverse tomorrow if the industry consolidates further and competition diminishes.

We're at an inflection point. The technology to democratize discovery exists. The pricing models to make it accessible are emerging. The question is whether the industry has the will to follow through -- or whether we'll find new ways to gatekeep access to the tools of justice.

I know which outcome I'm fighting for.