Snapshot
Continue to About ↓- What it does
- Nace.AI provides specialized small AI models for professional finance and accounting work: Drex, a decision model that returns a calibrated probability for each option instead of a free-text answer, and Nace Document Intelligence, which parses, classifies, splits, extracts and grounds documents with page-level citations.
More detail
- Enterprise teams and professional firms use it to automate document-heavy back-office and front-office workflows, optionally running inside their own environment.
- Best for
- Enterprise in-house finance teams automating AR/AP and accounting work, and audit or accounting firms that want an assistant that works directly on their own files.
More detail
- Also suited to developers and platform teams building document extraction, triage or confidence-gated review workflows through the APIs, SDKs, CLI or MCP Server.
- Pricing
- Usage-based rather than a monthly subscription: Nace's documentation describes a per-document allowance reserved at admission, cost varying by chosen effort tier (low, medium, high) and by whether parse results are reused.
More detail
- Work is refused (HTTP 402) while an account is paused for an unpaid invoice,
- Main limitation
- Buyers looking for a general-purpose generative chatbot or open-ended writing assistant, since Drex is designed to return options with probabilities rather than compose answers.
More detail
- Buyers who need guaranteed throughput on a prepaid balance, or who need independently benchmarked performance, should note that admission is capped by account balance and independent review coverage was not found.
- Record
- Living and monitored
- Reviewed
- Evidence
- Public web research
- Machine access
- Structured profile ↗
About
- Nace.AI provides specialized small AI models for professional finance and accounting work: Drex, a decision model that returns a calibrated probability for each option instead of a free-text answer, and Nace Document Intelligence, which parses, classifies, splits, extracts and grounds documents with page-level citations.
- Enterprise teams and professional firms use it to automate document-heavy back-office and front-office workflows, optionally running inside their own environment.
What it helps with
- General-purpose LLMs are hard to audit and can be confidently wrong; Nace returns calibrated probabilities per option an
- General models are blind to an organization's own policies; the Metamodel adapts a model to organization-specific polici
- Manual document intake and review: classify packets, split them into documents, and extract values into a JSON schema wi
- Automation that needs a confidence threshold: triage and routing can act only when the model is sure enough and send low
- Data sovereignty: models and data can run inside the customer's own perimeter via self-hosted, managed or hybrid deploym
- Fraud, credit and filings decisions via the newer Drex for Finance model.
Capabilities
A practical view of what is available and where limitations apply.
| Offering | Available | Detail |
|---|---|---|
| Drex decision model | Yes | Send state plus named questions (yes/no, multiple choice or scale) and receive probabilities for every option in one call; marketed as under 200 ms per decision. |
| Document Intelligence (Perception) | Yes | Five chainable tools at POST /v1/documents/{tool}: parse, extract, classify, split and ground, so a file is parsed once and reused. |
| Page-level citations | Yes | Extraction returns each field with status and a citation, and grounding locates each quote exactly in the source. |
| Drex for Search | Yes | Marketed on the homepage as the search component of the execution layer, finding the page inside a customer's files. |
| Self-hosted / in-perimeter deployment | Yes | Self-hosted (VPC, edge, on-premises), managed and hybrid options are listed on the official site. |
| Developer access methods | Yes | Console, APIs, CLI, SDKs, MCP Server for agent clients, and add-ins such as Nace for Excel. |
| Drex for Finance | Yes | Announced Nov 2, 2026 for fraud, credit and filings; the 77.3 vs 63.6 index comparison is a first-party benchmark claim. |
| SOC 2 Type II | Yes | The company states it is SOC 2 Type II audited, with the report and subprocessors at its trust center. |
| Independent review validation | Unknown | No independent review coverage of Nace.AI specifically was found in the sources retrieved. |
| Free tier | Unknown | No free plan or trial terms were disclosed on the public pages retrieved; pricing is usage-based and admission is capped by balance. |
Bot trust
How confidently AI systems can verify this profile from public evidence.
—of 100
ChatGPT—/100
Claude—/100
This profile has not completed a Bot Score assessment yet.
Evidence
Where this record came from, when it was reviewed, and how confidently it can be used.
- Verification
- Scribble researched these facts from the linked public sources.
- The owner has not manually confirmed every field.
- Last reviewed
- Primary source
- nace.ai
Technical profile details
Record ID
b29fc13a-e17d-4cfb-8773-d9721caedb2fCanonical URLhttps://profiles.scribble.network/listing/nace-ai-b29fc13a