Iris.ai
Enterprise AI harness for R&D documents and research
Overview
A Norwegian company (Iris.ai AS, registered 2015) that pivoted from an AI paper-finding assistant to an enterprise "AI harness": Axion turns PDFs, tables and charts into structured data, Neuralith runs checked AI agents in the customer's cloud or on premise, and RSpace offers concept search over papers and patents for R&D teams.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- One-time extraction of up to 10 documents
- Basic text and table extraction
- Quality report
- Up to about 100k pages with 2 million credits
- Advanced table and metadata extraction
- Technical document translation (68 languages)
- Forward-deployed engineers
- Unlimited data volume
- Governance and compliance layer
- LLM evaluation and monitoring
- In-house LLM integration
Use cases
What you can produce with Iris.ai
- PDFs, slide decks, tables and charts turned into structured, AI-ready data with Axion
- AI agents run in your own cloud or on premise with Neuralith
- Concept search over papers and patents in RSpace
- A document-extraction pilot on Axion Starter before a volume deal
- Volume extraction from €0.05 per page on Axion Enterprise
Iris.ai is no longer the scientific-search assistant it launched as in 2015. After what its co-founder calls a substantial commercial pivot, it sells an enterprise AI harness to manufacturing, life-sciences, energy and telecom teams: Axion extracts structured data from PDFs, tables and charts, Neuralith runs agents in a closed loop with checks on every answer, and RSpace covers literature and patent search for R&D.
Everything is demo-led; the only published price is an Axion Enterprise rate from €0.05 per page. The company is small, with about €16 million raised and losses in its filed 2025 accounts, so reference calls and a pilot matter more than marketing figures. Individual researchers should use free tools such as Semantic Scholar.
What we found
Iris.ai has reinvented itself more than once. It started in 2015 as what co-founder Anita Schjøll Abildgaard called "an AI Researcher", a tool for finding scientific papers. In a December 2025 post marking ten years, Abildgaard wrote that the company had made "a substantial commercial pivot" over the previous year. Iris.ai now calls itself an AI harness for expert knowledge and sells to manufacturing, life-sciences, energy and telecom teams that run it on their own documents.
There are three products. Axion pulls text, tables, charts and images out of PDFs, slide decks and scans in more than 68 languages and turns them into structured data that search tools and AI agents can use. Neuralith runs AI agents in a closed loop, in the customer's cloud or on premise, and the vendor says more than 23 checks gate every answer. RSpace is the literature tool for R&D teams: concept search over papers and patents with filters you control, which Iris.ai claims makes reviews 25–40% faster. That figure is the vendor's own.
Iris.ai AS is a Norwegian company registered in November 2015. Its Series A, announced in May 2024, raised €7.64 million and took total funding to about €16 million; an earlier version of this page gave a larger dollar total we could not trace. Its filed accounts show revenue of NOK 9.6 million for 2025 and a loss of NOK 26.2 million. The published case studies are ArcelorMittal, which uses Axion in its R&D process, an unnamed telecom group and a public-sector client.
Every product is sold through a demo. Axion is the only one with a public tier outline, and the only price it shows is an Enterprise volume rate from €0.05 per page. There is no free plan or self-serve sign-up. Researchers who just want to find papers should use a free tool such as Semantic Scholar. Iris.ai has no ASEAN office, and the three jobs it is advertising are remote.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
Notable facts
- Iris.ai was the first AI tool to generate a visual map of a scientific research landscape from a single seed paper, predating similar features in commercial tools by several years.
- The platform's extraction AI can read 1,000 clinical trial papers and populate a structured database with patient demographics, dosages, and outcomes in hours rather than weeks.
- Iris.ai was incubated at CERN, the European particle physics laboratory — one of the world's leading scientific research organisations.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including Iris.ai's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Iris.ai unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Iris.ai directly →
Spotted something out of date? Suggest an update →
Related reading
Health & Wellness
Fine-Particle Air Pollution Linked to Higher Hospital Admission Risk for Up to 21 Days, St...
Health & Wellness
A Guide for New Mothers Missed Its Main Goal but Nearly Doubled Primary Care Visits
Health & Wellness
Only 1 in 32 Young Americans With Opioid Addiction Stayed on Medication for Six Months
Alternatives to Iris.ai
More in Research & Data
