---
title: "Drex for Finance: fraud, credit and filing decisions for fintech"
description: "A decision model for fintech engineers: score fraud, credit, filings and support with answers limited to the options you define. 77.3 on the Drex Finance Index of 48 held-out finance test sets, against 63.6 for Jev 1.14, and ahead of it in all 9 categories."
canonical: "https://nace.ai/drex/finance"
updated: 2026-11-02
---

# Drex for Finance: fraud, credit and filing decisions for fintech

> A decision model for fintech engineers: score fraud, credit, filings and support with answers limited to the options you define. 77.3 on the Drex Finance Index of 48 held-out finance test sets, against 63.6 for Jev 1.14, and ahead of it in all 9 categories.

## Drex for finance

Fraud, credit and filing decisions for fintech. You list the options, Drex picks one and says how sure it is. It can't answer outside your list.

- 85.7% accuracy on 3,361 finance test decisions, vs 76.7% for Jev 1.14
- 9 of 9 benchmark categories ahead of Jev 1.14, from filings to credit risk

## What you can use it for

- Documents & filings. Example: Is this invoice a duplicate of one already paid? (Yes, No)
- Credit & lending. Example: Which risk band does this applicant fall in? (Low, Medium, High, Decline)
- Fraud & AML. Example: Which chargeback reason fits this dispute? (Fraud, Not received, Duplicate, Other)
- Markets & trading. Example: How does this Fed statement read? (Hawkish, Neutral, Dovish)
- Insurance & regtech. Example: Is this claim covered under the policy as written? (Covered, Not covered, Needs review)
- Banking & operations. Example: Which team should this banking request go to? (Cards, Loans, Payments, Fraud team)

## Benchmark results against Jev 1.14

The Drex Finance Index combines scores on 48 public finance test sets the model never saw in training. Drex for Finance scores higher than Jev 1.14 in all 9 categories.

| Benchmark | Drex for Finance | Jev 1.14 | GPT-6 Luna | pplx-decider v1.2 |
| --- | ---: | ---: | ---: | ---: |
| Drex Finance Index | 77.3 | 63.6 | 58.4 | 65.0 |
| Accuracy, % | 85.7 | 76.7 | 70.9 | 77.4 |
| Sentiment & tone | 80.7 | 69.1 | 68.6 | 73.9 |
| News & headlines | 89.5 | 66.4 | 60.7 | 68.8 |
| Banking & support intent | 87.1 | 71.6 | 71.9 | 73.7 |
| Documents & extraction | 92.3 | 85.9 | 48.8 | 87.1 |
| Financial QA & filings | 96.7 | 86.4 | 88.1 | 82.6 |
| Credit & risk | 48.3 | 30.0 | 38.3 | 11.7 |
| Monetary policy | 53.8 | 45.0 | 38.8 | 35.0 |
| ESG & climate | 73.0 | 59.1 | 48.1 | 54.4 |
| Markets & reasoning | 41.5 | 28.1 | 30.1 | 43.1 |

0 is a coin flip, 100 is every answer right. Bold is the best score in each row. The finance training didn't cost general skill: JevBench went from 93.9% to 94.4%. All seven models are compared in the announcement.

## How to get started

Install nace-sdk for Python or TypeScript and set the model to drex-max-finance. Already on Jev? Point the TypeSafe SDK at the Console and change the model name.

```python
# pip install nace-sdk
from nace_sdk import NaceClient, Noul

with NaceClient(model="drex-max-finance") as client:
    result = client.system_one(
        state="INV-2304, Northwind Supply, $9,870.00. "
              "Paid Oct 12: INV-2304A, Northwind Supply, $9,870.00.",
        questions={
            "duplicate": Noul(instructions="Is this invoice a duplicate?"),
        },
    )

print(result.nouls["duplicate"].noul)
```

- [Open Console](https://console.nace.ai?utm_campaign=drex-finance-launch&utm_content=finance&utm_medium=website&utm_source=nace.ai)
- [Read the announcement](https://nace.ai/blog/introducing-drex-for-finance)
- [API docs](https://console.nace.ai/docs)
- [MCP server](https://console.nace.ai/docs/developer-tools/mcp)

[Use cases](https://nace.ai/use-cases)

## 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/perception.md): the Perception series: parse, split, classify, extract and ground any file type with a location on every value
- [Drex for Search](https://nace.ai/drex/search.md): Drex for Search: decision-powered retrieval for agents, Drex finds the page, cited search through the CLI, SDKs, MCP and REST, and the public benchmark ranks
- [Use cases](https://nace.ai/use-cases.md): every Drex use case with its proof figures: browser use, search and finance for developers, enterprise finance teams and professional firms
- [Developers](https://nace.ai/developers.md): The execution layer for agents: Perception, Decision and Search behind one key, the Drex Max frontier, install steps, a scripted agent session and the four deployment boundaries
- [Site index](https://nace.ai/llms.txt): every page on nace.ai; append `.md` to any page URL or send `Accept: text/markdown` for the markdown version