For law firms
A pre-configured box, shipped to your office. Plug it in — you're operational.
~70% of requests processed locally. ~30% routed to the best cloud LLMs (Claude, GPT, Mistral): only anonymized excerpts. Leased for 24 months, then yours: the box and the software.


Extraction, Q&A, summaries, drafting
Anonymized before sending
The Problem
Consumer AI assistants and per-seat legal SaaS: your queries pass through third-party cloud servers. Client data leaves your office, potentially waiving privilege.
Every tool is locked to its model vendor. The best model changes every month: your tool doesn't.
ABA Model Rules and state bar ethics opinions increasingly scrutinize AI use. Most legal AI tools weren't built with privilege preservation in mind.
The Solution
A physical box shipped to your office, pre-loaded with the RAGbase software. Plug it in — you're operational.
Local models for everyday tasks. Cloud LLMs (Claude, GPT, Mistral) for complex queries — only anonymized excerpts.
Hardware provided and pre-configured. No infrastructure to manage. No IT department needed. Plug and play.
Model-agnostic. When a better model releases, we deploy it on your box. Your system stays the same.
Firm documents
Emails, DMS, contracts, briefs
RAGbase Box
At your office
~70% Local
Extraction, Q&A, summaries
~30% Cloud LLMs
Complex reasoning
Only anonymized excerpts are sent to cloud LLMs. Full documents never leave the box.
In practice
From client intake to final pleading, every task is assigned to the right sovereignty layer.
30 / 40
tasks stay in Europe. US models are an option, never a requirement.
nothing leaves the laptop
at the firm, local models
Mistral, anonymized excerpts
Claude / GPT, you decide
By default, nothing leaves the EU. The right-hand column is enabled matter by matter, and can be switched off. Hover a task for details.
Compliance
As state bar ethics opinions and ABA guidance on AI tighten, the RAGbase Box is architected for compliance by default.
Client documents stay on the box at your office. 70% of tasks are processed locally — no external connection.
Built-in de-identification before any data reaches cloud LLMs. Only anonymized excerpts are transmitted.
Full client documents never leave the firm's premises. Reasoning traces and agent logs stay on the box.
Every query, every response, every source is logged. Full audit trail on the box for compliance review.
Neither local models nor cloud providers use your data for training. Contractually guaranteed via enterprise APIs.
Architecture designed for ABA Model Rules 1.1, 1.6, 5.3. Documentation provided for your firm's compliance review.
Plans
One monthly rate for the whole firm: box, software, installation, updates, and support included. No per-user pricing, no upfront cost. After 24 months, the box and the software license belong to the firm.
Solo / small firm · 1 to 10 attorneys
24 months all-inclusive. At term, the box and license are yours.
Hardware included
Mac mini M4 Pro: 48 GB RAM, 1 TB SSD, encrypted backup drive, UPS
Qwen3 30B (MoE) by default, Mistral Small 24B as an alternative
Established firm · 10 to 40 attorneys
24 months all-inclusive. At term, the box and license are yours.
Hardware included
Mac Studio M3 Ultra: 96 GB RAM, 2 TB SSD, encrypted backup drive, UPS
100 to 120B-class models (gpt-oss-120b, Mistral Small 4): faster, sharper
40+ attorneys, groups, multi-office
Sized to your volumes. Same principle: yours at term.
Hardware included
Multi-GPU Linux station or box cluster, sized to your volumes
Models of your choice, calibrated to your use cases, 20 to 50 concurrent inferences
A 24-month all-inclusive plan: hardware, software license, installation, updates, monitoring, and support. At term, the box and the license belong to the firm. Details below.
Monthly plan
RAGbase software license (agents, RAG, traceability engine), transferred to the firm at term
Local model updates as soon as they release
Intelligent routing to cloud LLMs (~30%): Claude, GPT, or Mistral based on task; anonymized excerpts only
Box monitoring, security patches, encrypted backups
Hardware included: maintenance, loaner unit on failure, ownership transferred at the end of the 24 months
Technical support (level depends on plan)
Calibration on your firm's drafting styles and templates
Billed monthly per firm, 24-month term, then ownership transfer. Afterwards, optional Continuity program (updates, monitoring, support): $390/mo for Box Solo, $650/mo for Box Firm.
Comparison
| RAGbase Box | Per-seat legal SaaS | General AI assistant | |
|---|---|---|---|
| Where your data runs | Box at your office (local); only anonymized excerpts reach the cloud | Vendor cloud, full documents | Vendor cloud, full queries |
| Data privacy | ~70% local + anonymized cloud via enterprise APIs | Full documents in vendor cloud | Full queries in vendor cloud |
| AI model | Model-agnostic: always the best model | Locked to one vendor | Locked to one vendor |
| Pricing | Flat all-inclusive firm plan; box and license owned at term | Per seat, software only, forever | Per seat, forever |
| Your internal docs | 15+ years indexed, stay at the firm | Manual upload, external cloud | No permanent indexing |
| Attorney-client privilege | Full documents never leave the box | Third-party cloud access | Third-party cloud access |
| If you leave | The box, your data, and the index stay with you | Everything stops | Everything stops |
Frequently asked questions
Routine tasks (search, cited Q&A, summaries, extraction, filing) run directly on the box at your office with no external connection. The most demanding requests (long-form drafting, complex reasoning) can be routed to a cloud LLM: only anonymized excerpts are transmitted, never your full documents. A 100% local mode is also available: one switch, global or matter by matter.
Full client documents never leave your office. The RAGbase Box processes ~70% of tasks locally. For requests routed to cloud LLMs, only anonymized excerpts are sent via enterprise APIs that contractually guarantee zero retention and zero training on your data. This architecture was designed to preserve privilege by default.
Both, in order. For 24 months you pay one flat monthly rate that covers everything: hardware, software, installation, updates, monitoring, support. At the end of the term, the box and the software license are transferred to the firm: they belong to you. This is not an endless rental; it is a purchase spread over 24 months.
The box is yours and keeps working, with your data and your index, even if you stop paying anything. To keep model updates, security patches, monitoring, and support, the optional Continuity program is available: $390/mo for Box Solo, $650/mo for Box Firm, no commitment. You can also trade in your box for the next hardware generation.
For routine tasks (~70% of volume), local models deliver cited answers in seconds, on-premise. For very long legal reasoning over a large matter, frontier cloud models keep an edge: that is exactly what the hybrid routing covers, sending only anonymized excerpts.
The RAGbase Box is not locked to any model vendor. When a better local model releases (Qwen, Mistral, gpt-oss...), we deploy it on your box as part of updates. For cloud, routing picks the model suited to each task. Your interface and workflows stay the same: only the engine improves.
Text generated locally on the box carries no vendor watermark: your firm sets its own policy. Since August 2026, under the EU AI Act, major cloud providers embed invisible marks in generated text (Anthropic announced it on August 11, 2026). This only concerns requests routed to the cloud, and that routing stays under your control, matter by matter.
The RAGbase Box architecture is designed for compliance with ABA Model Rules and state bar ethics opinions on AI use. Client documents stay on-premise, cloud data is anonymized, and full audit trails are maintained. We provide documentation to support your firm's compliance review.
A loaner unit ships while yours is repaired, and your data restores from the encrypted backup drive that stays at your office. Back up and running within hours. Your data never passes through us.
Order
A simple three-step process. Operational in 2 to 4 weeks, installation and training included.
30 minutes
Book a 30-min call with our team. We analyze your needs, query volume, and constraints (number of attorneys, DMS, security requirements).
Quote in 24h
We size the right hardware, select the optimal AI models for your firm, and prepare a detailed quote.
Shipped to your office
The box ships pre-configured. Guided installation with our team, your archive indexes overnight, then 2 h of team training. Total lead time: 2 to 4 weeks depending on hardware stock.
Book your free 30-minute consultation. No commitment — we answer all your questions.
No commitment · Quote within 24h · Installed within 2 to 4 weeks
Book 30 minutes with our team. We analyze your needs, size the right box, and ship it to your office.
Free consultation · No commitment
We use audience and marketing cookies (Google Analytics, LinkedIn). No tracker loads without your consent. Learn more