# Nace.AI

> Execute your Workflows with Nace Small Models Powered by the Metamodel, our specialized models deliver exceptional precision at one-tenth the cost of large LLMs. Automate audit, accounting and billing with our applications, or build your own with our models and APIs.

## When to use Nace

- Make a decision from the facts of a case: send Drex a state and typed questions (pick one of several, yes or no, rate on a scale) and get a calibrated probability for every option in one forward pass, with no text to parse.
- Put a decision model inside a workflow: threshold the confident calls, route the uncertain ones to a person, and tune Drex on your own labelled outcomes.
- Read any file type with Nace Document Intelligence: parse, split, classify, extract and ground, with every value tied to the page and box, sheet and cells, or span it came from.
- Run the same models in our cloud, a dedicated cluster, your own cloud account or air-gapped, with your data training only your models.
- Automate receivables, payables, monitoring and continuous audit for a finance team, or put an AI Senior Accountant on an audit and accounting firm's clients.

## Get started for agents

- [Drex API guide](https://nace.ai/drex.md): the decision model: request shape, question types, benchmarks and deployment, as markdown
- [Decision endpoint](https://console.nace.ai/v1/systemone): POST with model "drex-latest", a state string and typed questions (choice, noul, score); https://console.nace.ai/v1/models lists checkpoints
- [Document endpoints](https://console.nace.ai/v1/documents/{parse|split|classify|extract|ground}): POST a source (a public URL, an uploaded workspace file or a finished parse job) and poll GET https://console.nace.ai/v1/documents/jobs/{id}; the same key and credit as decisions
- [Console](https://console.nace.ai): API keys, the playground, usage and billing; send `Authorization: Bearer $NACE_API_KEY` on every request
- [Console docs](https://console.nace.ai/docs): quickstart, question types, the document routes, errors, limits and pricing
- [Nace Document Intelligence](https://nace.ai/ndi.md): the Perception series: parse, split, classify, extract and ground any file type with a location on every value
- [Developers](https://nace.ai/developers.md): The execution layer for agents: Perception, Decision and Context behind one key, the Drex Max frontier, install steps, a scripted agent session and the four deployment boundaries

## How to read this site

- Any page as markdown: append `.md` to its URL or send `Accept: text/markdown`. Responses set `X-Markdown-Tokens` (an estimate) and link the HTML canonical.
- [Full site as one file](https://nace.ai/llms-full.txt): every listed page, serialized in order.
- [Sitemap](https://nace.ai/sitemap.xml)

## Business

- [Nace for business](https://nace.ai/business): Nace for business: in-house finance teams automate receivables, payables, monitoring and audit; audit and accounting firms run an AI Senior Accountant. Small models, trained on your work, inside your perimeter.
- [Nace for enterprise](https://nace.ai/business/enterprise): Nace for enterprise: build your own intelligence instead of renting it. Small language, decision and execution models that automate receivables and payables end to end, follow your policies, and run in your cloud or your data center.
- [Nace for professional firms](https://nace.ai/business/professional-firms): Nace for audit and accounting firms: an AI Senior Accountant that performs bookkeeping, control and reports, end to end, with every control run on every entry. Models the firm owns, inside the firm's tenant, running in 24 hours.

## Developers

- [The execution layer for agents](https://nace.ai/developers): Nace is the execution layer for agents: Perception reads any file, Decision makes the call in under 200 ms. One API key, at small-model speed and cost.
- [Nace Document Intelligence](https://nace.ai/ndi): Nace Document Intelligence parses, splits, classifies, extracts and grounds 36 file types with a location on every value. First on grounding and level with the frontier models on split and extract in Nace's Perception evaluation, managed, in your cloud or air-gapped.
- [Drex Max, the decision model that sees](https://nace.ai/drex): Drex Max decides from screens, images and video, built for agents that use computers and browsers. It scores 67.59 on the public Decision Index, 6.9 points over Jev 1.14, decides in under 200 ms and accepts every screenshot size. Drex 2.0 ships open weights, and Drex 1.5 routes text at volume.
- [Kapitsa, local inference for Apple silicon](https://nace.ai/kapitsa): Kapitsa is an open-source inference engine built for Qwen and Nemotron on Apple silicon. It writes 154 tokens per second on Qwen3.6 35B-A3B on an M4 Pro, ahead of oMLX, Ollama, LM Studio and llama.cpp, and serves OpenAI and Anthropic APIs on your Mac.

## Research

- [Research](https://nace.ai/research): Research from Nace.AI on training trustworthy models capable of executing tasks reliably while continuously self-improving.
- [Scaling Laws for Hypernetwork-Based Knowledge Injection](https://nace.ai/research/scaling-law-for-hypernetwork): New scaling laws for train-time knowledge injection: hypernetworks beat LoRA and full fine-tuning on out-of-distribution generalization, and the gap grows with target model size. (2026-07-22)
- [Enterprise Policy Injection with Metamodels](https://nace.ai/research/enterprise-policy-injection-with-metamodels): Adapting model behavior at inference time without retraining, large context windows, or touching the primary model's weights. (2026-05-01)
- [Nace Meta Agent](https://nace.ai/research/nacemetaagent): NEMA: meta-learning architectures that generate specialized, task-specific agents automatically. (2026-01-27)
- [Memory Augmentation and Editing Techniques in LLMs](https://nace.ai/research/memory-augmentation-and-editing-techniques-in-large-language-models): Memory augmentation and model editing for handling outdated and domain-specific knowledge. (2025-03-18)
- [NAVI Console: Policy-driven Safeguards Comparison](https://nace.ai/research/policy-driven-safeguards): Benchmarking Nace's Safety Model against AWS Bedrock, Azure, NeMo Guardrails, and prompt-based approaches. (2025-01-19)
- [GeLoRA: Efficient Fine-Tuning of Large Language Models](https://nace.ai/research/gelora): A theoretical framework connecting the geometry of data representations to training dynamics. (2024-12-10)
- [Hypernetworks for Specialized Instructions](https://nace.ai/research/hypernetworks-for-specialized-instructions): Efficient fine-tuning of LLMs through task-specific parameter generation. (2024-12-07)
- [Meta-Learning through Hypernetworks](https://nace.ai/research/meta-learning-through-hypernetworks): Improving task adaptability in large language models. (2024-12-07)
- [Memory Augmentation for Document-heavy Enterprise Use Cases](https://nace.ai/research/memory-augmentation-for-document-heavy-enterprises): A modular RAG pipeline tailored to enterprise needs: over 90% F1 on verification. (2024-11-05)
- [Knowledge Graph Synergy with LLMs](https://nace.ai/research/knowledge-graph-utilization): How structured, verifiable knowledge complements LLMs where they fall short. No ML background required. (2024-11-01)
- [Progressing Towards Meta Models and What Lies Beyond](https://nace.ai/research/progressing-towards-meta-models-and-what-lies-beyond): Tailored, meta-learned Small Language Models for rapid, efficient adaptation. (2024-10-24)
- [Understand LLMs From the Perspective of Ranking](https://nace.ai/research/understand-llms-from-the-perspective-of-ranking): Most post-training performance gains are reranking: aligning the proposal distribution with the task. (2024-10-12)

## Blog

- [Blog](https://nace.ai/blog): Announcements, product launches and company updates from the Nace.AI team.
- [Introducing Drex Max, a multimodal decision model for computer use](https://nace.ai/blog/introducing-drex-max): Drex Max is a decision model that sees the screen. It reads text, images and video, ranks first among decision models on the public Decision Index 0.3.2, and is generally available today. (2026-10-19)
- [Introducing Nace Document Intelligence: The Perception Layer for AI Agents](https://nace.ai/blog/document-intelligence-launch): Five tools that parse, split, classify, extract and ground any of 36 file types, and point every value back to the page it came from. Ranks first on grounding against frontier models. Available in the Console today. (2026-10-08)
- [Introducing Drex, a small model that decides instead of writing](https://nace.ai/blog/introducing-drex): Drex returns a calibrated probability for every option in one forward pass, and leads the public Decision Index with a model small enough to serve on one accelerator. (2026-09-24)
- [Nace.AI raises $21.5M to build AI-led enterprise workflows](https://nace.ai/blog/announcing-our-seed-investment): Nace.AI raises $21.5M to build specialized intelligence for complex enterprise workflows. (2026-05-05)
- [A Letter from the Founders of Nace.AI](https://nace.ai/blog/letter-from-founders): Nace.AI emerges from stealth with Metamodel 1: precision and adaptability for enterprises. (2025-02-08)

## Company

- [About Us](https://nace.ai/about): Nace.AI builds small models for professional work. Founded in 2024 by researchers and engineers from Google, Meta, Amazon and the University of Toronto, headquartered in Palo Alto.
- [Careers](https://nace.ai/careers): Join Nace.AI. We build small models for professional work in Palo Alto, CA. See our open positions.
- [Contact Sales](https://nace.ai/contact): Schedule a call with the Nace.AI team. Tell us about your organization and what you are looking to do, and we will get back to you.
- [Press](https://nace.ai/press): News coverage of Nace.AI, company announcements, media contact, and downloadable logos and team photos for editorial use.
- [Privacy Policy](https://nace.ai/policies/privacy-policy): How Nace.AI collects, uses and protects your personal information across our websites, products and services.
- [Terms of Service](https://nace.ai/policies/terms-of-service): The terms and conditions governing your use of Nace.AI websites, products and services.

## Community and trust

- [Trust center](https://trust.nace.ai): SOC 2 Type II report, penetration test report and subprocessors
- [Discord](https://discord.com/invite/nace-ai): community server: questions, product updates, early access
- [X](https://x.com/NaceAI): @NaceAI
- [LinkedIn](https://www.linkedin.com/company/nace-ai/): company page
- [GitHub](https://github.com/nace-ai): open models, agent skills and tools

## Optional

- [Full site](https://nace.ai/llms-full.txt): every listed page as one markdown file
- [Developers index](https://nace.ai/developers/llms.txt): the model series, the Drex and Perception pages and deployment
- [Products index](https://nace.ai/products/llms.txt): product pages only
- [Research index](https://nace.ai/research/llms.txt): publications, newest first