Iris.ai vs Lexis+ with Protégé (2026)

Iris.ai

by Iris.ai

Lexis+ with Protégé

by LexisNexis

Pros

• ✓Covers the path from raw documents to governed knowledge for AI agents, with Axion for data preparation, Neuralith for knowledge graphs and RSpace for research.
• ✓Named customers span manufacturing, agriculture, publishing, research and government, including Mercedes-Benz, ArcelorMittal, Syngenta, USDA and the Max Planck Society.
• ✓Deploys through an API, on-premise or in the cloud, and states ISO 27001 certification and that client data is never used to train general models.

Limitations

• –Axion and Neuralith are sold through a demo request. Only the RSpace Individual Membership has a published price, at EUR 292 per month or EUR 3,500 per year.
• –No public API documentation or MCP server is published, and Axion and Neuralith deployments are scoped with Iris.ai's team rather than set up self-serve.
• –RSpace's cloud import accepts English-language documents only, up to 10,000 documents and 5 GB, and its trial moves to a paid period automatically after 10 days.

Pros

• ✓Outputs are grounded in LexisNexis's legal content collection with Shepard's Citations validation, so citations come with status and treatment signals rather than potential AI hallucinations.
• ✓Protégé Work orchestrates multi-step legal tasks through structured plans with visibility and control at each stage, and hundreds of pre-built workflows plus a no-code Custom Workflow Builder for firm-specific playbooks.
• ✓Integrates across Microsoft 365 (Word, Excel, PowerPoint, OneNote, Teams, Outlook), iManage, SharePoint, OpenText and Google Drive, with OpenAI and Anthropic as named LLM partners, so AI works inside existing legal workflows rather than requiring tool switching.

Limitations

• –Pricing is quoted per organization and varies with size, capabilities and content access, so a buyer needs a LexisNexis quote. The Forrester TEI studies, commissioned by LexisNexis, model ROI but publish no subscription price.
• –Several capabilities are scoped by region or still to come: Mistral models are specific to the France edition, Arabic and English case translation to the Middle East edition, LexisNexis models are due for preview later in 2026, and customer-created skills are coming.
• –The G2 profile still carries the former Lexis+ AI name and holds only a handful of reviews since the February 2026 rebrand.