Home / Directory / Agents & orchestration / Agent platforms & frameworks / Nomic AI
Nomic AI
Nomic AI is a software company whose AI agents help architecture, engineering, and construction firms review drawings against building codes and firm standards, check submittals, research RFIs, and search project files.
Engineering consultancies, design firms, and contractors with large drawing sets, specifications, submittals, and RFIs, that want an agent to flag issues and cite the sheet or code section while an engineer stays accountable for the result.
Companies outside architecture, engineering, and construction, and anyone who wants findings they can accept without checking.
Verdict
Nomic is worth a look because large consultancies are rolling it out: Arcadis reports nearly 500 licensed users and more than 750 workflows, and Aurecon is scaling access across its 6,700 people. Both firms became Nomic investors in September 2026. The trade-offs are the newness of the focus and Nomic's own documentation, which calls a drawing review a first-pass aid whose rows need checking.
Score Breakdown
These scores cover how Nomic AI handles drawing and specification review, AEC systems integration, cited and checkable findings, and enterprise rollout and security, and then how it does as a company on innovation, communication, pricing, and contract terms.
Who Uses Nomic AI, Where It Fits on Price, and Its Competitors
Nomic AI is an agent platform for the people who design, engineer, and build physical projects. The agents read drawings, specifications, models, submittals, and RFIs in the systems that hold them, keep the permissions set in those systems, and cite the sheet, spec section, or code clause behind each finding. Pursuit teams, design reviewers, QA/QC reviewers, and contractors use them for drawing review against firm standards, code checks against more than 380 codes and standards, submittal logs, and questions across a project's files.
Nomic narrowed its focus to the built world in late 2025. It started in New York in 2022 with open-source data tools and models, including Atlas, GPT4All, and the Nomic Embed models, and announced the change in a November 3, 2025 post. Its CEO, Andriy Mulyar, told AEC Magazine in October 2026 that revenue in the earlier business had flattened because foundation-model companies absorbed what Nomic offered.
On price, Free is limited, Individual costs $20 a month, Business costs $40 per user a month with a 25-seat minimum and $20 of AI usage per seat, and Enterprise is custom. Among the nine agent platform vendors on our chart, only LangChain has a lower typical engagement cost, although the 25-seat Business minimum means a team plan starts at $1,000 a month at list price. AEC Magazine lists Autodesk, Procore, Nemetschek's Bluebeam, and Togal as competitors in this work, and Nomic itself lists general-purpose AI chat tools, AI bundled inside one authoring or project-management suite, and manual review.
Pricing
We checked Nomic's pricing page and recorded four plans. AI usage is a dollar balance that pays for agent chats, workflows, document indexing, and Agent API calls at the underlying model rates, with no markup on those model costs.
| Plan | Price | What stands out |
|---|---|---|
| Free | $0 | No credit card, limited agent requests, and access to core platform surfaces. |
| Individual | $20 / month | Extended agent requests, every product surface except the Agent API, included AI usage with on-demand top-ups, and a solo workspace. |
| Business | $40 / user / month, 25-seat minimum | $20 of AI usage per seat, project delivery, file integrations including Autodesk and Bentley, SAML/OIDC SSO, usage analytics, and Agent API access. |
| Enterprise | Custom | Pooled usage commits, invoice or purchase-order billing, SCIM, audit logs, hosting in the EU, Australia, US, and UK, priority support, and a dedicated customer success manager. |
The Field at a Glance
How Nomic AI compares with other agent platform and framework vendors we reviewed, by Overall Score and relative typical engagement cost.
Nomic AI scores 7.4 in agent platforms & frameworks, under LangChain (8.0), Gumloop (7.8), Mastra (7.6), and Kore.ai (7.5), and ahead of Photon (7.3), Runable (7.1), Dust (6.8), and Relevance AI (6.7).
How Well Nomic AI Fits Each Design and Construction Task
We rated Nomic AI against the tasks design and construction teams bring to an AI reviewer, drawing on its product and documentation pages, its pricing, and AEC Magazine's October 2026 profile.
| Use case | Fit | Notes |
|---|---|---|
| Drawing review against firm standards | Strong | Agents check drawing sets against QA/QC standards and design criteria and compare 60%, 90%, and issued-for-construction sets. Nomic's pricing FAQ puts $20 of usage at roughly 20 to 80 sheet reviews. |
| Code compliance checks | Strong | Drawings are cross-referenced against more than 380 building codes and standards, with the section cited for each finding. |
| Submittals and RFIs | Strong | Agents build submittal logs from the project manual and research RFIs across the project record. Nomic's FAQ puts $20 of usage at roughly 10 to 25 submittal reviews. |
| Searching a firm's project knowledge | Strong | Arcadis's Freek Matheij describes a knowledge database built on projects, and Nomic's homepage lists project knowledge search among its out-of-the-box agents. |
| Large-firm rollout | Strong | Nomic reports about 500 Arcadis engineers in 17 countries, and Aurecon is scaling access across its 6,700-person workforce. |
| Connecting to project systems | Mixed | Nomic lists Autodesk Forma, Bentley ProjectWise, Trimble ProjectSight, SharePoint, and Egnyte, and AEC Magazine describes the Bentley integration as in progress. |
| RFP response drafting | Mixed | Nomic's about page lists drafting response material from a firm's past projects, and the customer material centers on drawing review. |
| Output you can accept as is | Mixed | Nomic's documentation calls each findings row something to verify, and says a drawing review is a first-pass aid. |
| Small firms | Mixed | Business starts at 25 seats, which makes $1,000 a month at list price, and Individual covers one person. |
Review Excerpts
These excerpts come from AEC Magazine's October 2026 profile, Aurecon's quote on Nomic's pricing page, and Nomic's own documentation. Arcadis and Aurecon became Nomic investors in September 2026. Paid reviews and pay-for-play sites were excluded.
“It's kind of a knowledge database based on projects, and then you can have the right context, and that makes it very cheap to ask questions.”
“Nomic is maybe not the most sexy AEC platform out there, but it really helps in terms of everything in the asset life cycles.”
“Nomic empowers us to build and deploy domain-specific AI to our whole global design team.”
“Every finding comes with a citation, but each row still needs an engineer to check it before it goes into the project record.”
Where customer reviews couldn't be found, our analysts attempted to give a voice to buyer sentiment, drawing on real information from review sites, forums, and product documentation.
Based on: Nomic docs and AEC Magazine
Company Facts
- Founded
- 2022, in New York City (Nomic)
- Headquarters
- New York City, as a hybrid company (Nomic)
- Leadership
- Andriy Mulyar, founder and CEO, and Ben Schmidt, CTO (Nomic)
- Funding
- A $17M Series A led by Coatue in 2023, per Nomic's about page, and Aurecon and Arcadis became strategic investors on September 3, 2026 (Nomic about; Nomic news)
- Change of focus
- A November 3, 2025 post announced the focus on the built world, after earlier products Atlas, GPT4All, and the Nomic Embed models (Nomic news; Nomic about)
- Security
- SOC 2 Type II with annual independent audits, zero-data-retention agreements with model providers, and Enterprise hosting in the EU, Australia, US, and UK (Nomic security; Nomic about)
- Website
- nomic.ai
Key Leadership
Andriy Mulyar
Founder and CEO
His about-page biography says he left a machine learning PhD at NYU Courant to found Nomic in 2022, was an early engineer at Rad AI, led engineering as CTO, and became CEO in 2024 (Nomic).
Ben Schmidt
CTO
He left the NYU faculty in 2022 to join Nomic's founding team and created Deepscatter, a browser-based library for visualizing billions of points (Nomic).
Methodology
We scored Nomic AI on eight criteria, from drawing and specification review to contract fairness, and the Overall Score is their simple average on a 0 to 10 scale. The scores draw on reviews from public review sites and forums, with paid reviews and pay-for-play sites excluded, plus a hands-on look at the product, its pricing, and its documentation and training. Scores can change as we add reviews and product checks.