> For the complete documentation index, see [llms.txt](https://docs.viesus.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.viesus.com/discover/choose-your-interface.md).

# Choose Your Interface

All VIESUS interfaces wrap the same enhancement engine and produce identical output for the same input and settings. The right choice depends on where your workflow runs, what you're processing, and how much infrastructure you want to manage.

{% hint style="success" %}
**Try on Cloud, run production on-premise.** VIESUS Cloud is the fastest way to evaluate enhancement quality. Once you're happy with the results, on-premise is our recommendation for production — it runs locally, which is faster and more cost-efficient at scale. See [Cloud vs on-premise](#cloud-vs-on-premise) below.
{% endhint %}

<button type="button" class="button secondary" data-action="ask" data-query="Based on my use case, which VIESUS interface should I use — CLI, PDF Enhancer, Node.js, C/C++ SDK, or Cloud? Ask me what I need." data-icon="gitbook-assistant">Help me choose an interface</button>

***

## Quick decision

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-cover data-type="image"></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><h4>VIESUS Cloud</h4></td><td>No server infrastructure, pay per image. Also the easiest way to try AI upscaling without a local NVIDIA GPU.</td><td><a href="/files/E4JjOos7P7724jjCYkyn">/files/E4JjOos7P7724jjCYkyn</a></td><td><a href="/pages/R7yP5JYM8pP1CRGO9fBF">/pages/R7yP5JYM8pP1CRGO9fBF</a></td></tr><tr><td><h4>VIESUS CLI</h4></td><td>Batch-process images from the command line or a shell script.</td><td><a href="/files/E4JjOos7P7724jjCYkyn">/files/E4JjOos7P7724jjCYkyn</a></td><td><a href="/pages/jCaCCglTDTUi2XjWaZ4w">/pages/jCaCCglTDTUi2XjWaZ4w</a></td></tr><tr><td><h4>VIESUS PDF Enhancer</h4></td><td>Process PDF documents with embedded images — photobooks, catalogues, marketing print.</td><td><a href="/files/E4JjOos7P7724jjCYkyn">/files/E4JjOos7P7724jjCYkyn</a></td><td><a href="/pages/sKGMxwmEUyCd9vHKTvem">/pages/sKGMxwmEUyCd9vHKTvem</a></td></tr><tr><td><h4>VIESUS Node.js Module</h4></td><td>Integrate enhancement into a Node.js web server or cloud pipeline (Linux only).</td><td><a href="/files/E4JjOos7P7724jjCYkyn">/files/E4JjOos7P7724jjCYkyn</a></td><td><a href="/pages/4V13iUcemnYE7NIgOrfP">/pages/4V13iUcemnYE7NIgOrfP</a></td></tr><tr><td><h4>VIESUS C/C++ SDK</h4></td><td>Embed enhancement inside a C/C++ application without a subprocess call.</td><td><a href="/files/E4JjOos7P7724jjCYkyn">/files/E4JjOos7P7724jjCYkyn</a></td><td><a href="/pages/MohKFreJOs5UlHaCwPkg">/pages/MohKFreJOs5UlHaCwPkg</a></td></tr><tr><td><h4>Desktop Tools</h4></td><td>Windows GUI apps — tune enhancement settings interactively with the <strong>VIESUS Viewer</strong>, or run automated hotfolder pipelines with the <strong>Folder Enhancer</strong>.</td><td><a href="/files/E4JjOos7P7724jjCYkyn">/files/E4JjOos7P7724jjCYkyn</a></td><td><a href="/pages/svG0OJsD62p7ulF0rXUu">/pages/svG0OJsD62p7ulF0rXUu</a></td></tr></tbody></table>

***

## Cloud vs on-premise

Every interface produces the same enhancement, so the real decision is **where it runs**. VIESUS Cloud is the easiest way to **try** enhancement — no install, no GPU, and a free tier to evaluate. Once you're satisfied with the quality, we recommend running **on-premise for production**: it runs locally, which is roughly **twice as fast** and — since you're not paying for hosted infrastructure on every image — **steadily cheaper the more you process**.

How the costs compare depends on whether you need AI upscaling and at what volume:

<table><thead><tr><th width="255.800048828125">Option</th><th width="136">Up-front cost</th><th width="205.39990234375">Per AI-upscaled image</th><th>Break-even</th></tr></thead><tbody><tr><td>VIESUS Cloud</td><td>None</td><td>~CHF 0.06 (4 credits; CHF 75 = 5,000 credits)</td><td>—</td></tr><tr><td>On-premise — no AI upscaling</td><td>None</td><td>—</td><td>Immediately</td></tr><tr><td>On-premise — GPU upgrade (RTX 4000 Ada)</td><td>~CHF 1,500</td><td>None</td><td>~25,000 images</td></tr><tr><td>On-premise — dedicated system (NVIDIA DGX Spark)</td><td>~CHF 4,000</td><td>None</td><td>~70,000 images</td></tr></tbody></table>

*Indicative figures, as of 06/2026. Without AI upscaling, on-premise has no per-image cost and needs no special hardware, so it's cheaper than Cloud from the first image. With AI upscaling, past the break-even on-premise keeps getting cheaper while Cloud keeps charging per image — adding a GPU to an existing workstation is the lowest-cost entry; a dedicated system suits higher sustained volume.*

**Rule of thumb:** evaluate on Cloud, run high volume on-premise.

***

## Detailed comparison

<details>

<summary>VIESUS CLI</summary>

**Best for:** Automated batch processing of image files on Windows or Linux servers.

The `viesus` command-line tool reads a list of image paths (or individual files), applies enhancement, and writes output to a configurable destination. It is designed to be called from shell scripts, cron jobs, or orchestration tools.

```bash
viesus -l images.lst -s -p config.json
```

**Strengths:**

* Simple to integrate into any pipeline — no programming required
* Scales by running multiple instances in parallel
* Works on Windows and Linux (x64, arm64)
* Supports Docker

**Limitations:**

* File-based I/O only — images must be written to disk before and after processing

→ [VIESUS CLI documentation](/reference/cli-reference.md)

</details>

<details>

<summary>VIESUS PDF Enhancer</summary>

**Best for:** Processing PDFs with embedded images — photobooks, marketing print, catalogs.

The `viesusPDF` tool opens a PDF, locates embedded images, enhances each one, and writes a new PDF with the improved images. It supports hotfolder monitoring mode (watches a folder for new PDFs) and standalone mode (processes a specific file or folder).

```bash
viesusPDF input.pdf output/ config.json
```

**Strengths:**

* Operates directly on PDFs without pre-extracting images
* Hotfolder mode for fully automated print workflows
* Generates XML processing reports per PDF

**Limitations:**

* Processes images embedded in PDFs; not suitable for standalone image files
* Not available as a Node.js module

→ [VIESUS PDF Enhancer documentation](/reference/pdf-cli.md)

</details>

<details>

<summary>VIESUS Node.js Module</summary>

**Best for:** JavaScript-based web servers, REST APIs, SaaS platforms where enhancement happens on upload.

The Node.js module is a native C++ addon that calls the VIESUS library in-process from Node.js. It uses a worker thread pool so enhancement runs off the main event loop without blocking.

```js
const viesusObj = new viesus.MyViesusObject(guid);
const result = viesusObj.Enhance(fromPath, toPath, iniPath, resPath);
```

**Strengths:**

* In-process — no subprocess overhead
* Worker thread pool keeps the event loop unblocked
* Scales to multiple CPUs or GPUs by adjusting pool size

**Limitations:**

* Linux only (Ubuntu 22.04+)
* Requires CUDA 12.6 driver for GPU features

→ [VIESUS Node.js Module documentation](/reference/node.js-module/overview.md)

</details>

<details>

<summary>VIESUS C/C++ SDK</summary>

**Best for:** Existing C or C++ applications that need enhancement as a function call, not an external process.

The SDK exposes a C API via `IsEnhance.h`. Link against the VIESUS shared library and call enhancement directly from your code. Any language with a C FFI can wrap it.

**Strengths:**

* No subprocess; minimal latency
* Access to per-image analysis data (detected faces, scene type, correction strengths)
* Any language with C FFI bindings can use it

**Limitations:**

* Requires C/C++ integration effort
* Access requires contacting Viesus AG

→ [VIESUS C/C++ SDK documentation](/reference/overview.md)

</details>

<details>

<summary>VIESUS Cloud</summary>

**Best for:** No-infrastructure integrations, testing AI upscaling, or pay-per-image billing.

VIESUS Cloud is a GraphQL API hosted by Viesus AG. Upload an image or PDF, trigger enhancement, and download the result. No library installation, no GUID, no server to manage.

```bash
curl -X POST https://api.viesus.cloud/graphql \
  -H "x-api-key: your-key" \
  -d '{"query": "mutation { ... }"}'
```

**Strengths:**

* No hardware or software setup
* Free tier with 200 credits to evaluate (no credit card required)
* AI upscaling without needing a local NVIDIA GPU
* Credit-based billing — pay only for what you process

**Limitations:**

* Requires internet access and file upload
* Not suitable for air-gapped environments
* Processing is subject to API rate limits and server availability

→ [VIESUS Cloud documentation](/reference/cloud-api/overview.md)

</details>

***

## Platform support matrix

| Interface      | Windows x64 |  Linux x64 | Linux arm64 |    macOS   |
| -------------- | :---------: | :--------: | :---------: | :--------: |
| CLI            |      ✓      |      ✓     |      ✓      |      —     |
| PDF Enhancer   |      ✓      |      ✓     |      ✓      |      —     |
| Node.js Module |      —      |      ✓     |      ✓      |      —     |
| C/C++ SDK      |      ✓      |      ✓     |      ✓      |      —     |
| Cloud API      |  ✓ (client) | ✓ (client) |  ✓ (client) | ✓ (client) |


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.viesus.com/discover/choose-your-interface.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.