FAQ

About CNAPS

What is CNAPS Studio?

CNAPS Studio is a no-code AI platform for image, video, and audio work. Start on Home, where each card is a complete AI task — background removal, upscaling, poster generation, transcription — that runs from a single input with no setup. When you want to change how a task works, open that same card on the canvas and rewire it yourself. CNAPS Studio includes 80+ AI models and 80+ processing tools, and every plan gives you access to all of them. Available as a cloud platform, and as on-premise or edge device deployments for enterprise.

What AI capabilities are available?

Access a curated library of open-source and external AI models across multiple categories:

• Image generation, editing, and enhancement
• Object detection and segmentation
• Background removal and replacement
• Image upscaling and restoration
• ControlNet structure-locked generation
• Multimodal language understanding
• OCR and document understanding
• Video generation and upscaling
• Speech and audio processing
• Text-to-speech with custom voice & VoiceDesign
• Text-to-sound-effects (SFX)
• Lip-sync (audio to talking-head video)

Where do the AI models come from?

CNAPS Studio bundles open-source models from global leaders — Meta (SAM2/SAM3/SAM3.1), Alibaba (QWEN), Google (Gemma), Black Forest Labs (FLUX), Z.ai (GLM), ByteDance (SeedVR2), MOSS, LTX, ACE-Step, BiRefNet — alongside external AI integrations via API: OpenAI (ChatGPT, GPT Image), Anthropic (Claude), and Google (Gemini, VEO, Nano Banana). Browse the full catalog in the AI Model Hub.

Do I need any coding experience?

Not at all. Everything on Home runs without building anything. On the canvas, you connect the processing steps you need visually. No Photoshop skills, no ComfyUI complexity, no command line required.

Home & Cards

What is Home?

Home is the first screen you see at cnaps.ai — signed in or not. It's a gallery of cards, and each card is a complete AI task that has already been built and tested for you — no canvas, no connections, no setup. Pick a card, swap in your own input, and run it. Common tasks include video super resolution, text-to-4K image generation, image editing, background removal, privacy blurring, virtual try-on, and OCR.

Who is Home designed for?

Marketers, operations teams, and anyone who needs AI results without learning a workflow builder. Existing Studio users also use Home for faster access to single-step tasks.

Do I need to build a workflow to use CNAPS Studio?

No. Everything on Home runs without building anything — pick a card, swap in your input, and run it. The canvas is there for when you want to change how a task works, not before. Any card can be opened on the canvas with one click, so you can start simple and go deeper whenever you want.

Does looking at a card's example use up my capacity?

No — and running it doesn't either, as long as you don't change anything. Every card opens with a finished example already on screen, which is a saved result rather than a fresh run. Click Run again without touching the inputs and the card runs that sample for free and shows you the result — text cards included. Capacity is only consumed once you swap in your own input.

How do I find the right card?

Use the search bar at the top of Home and describe what you want to do — "make my video sharper", "remove the background", "poster of me". New cards are added continuously, so search is usually faster than scrolling.

Can I see how a card actually works?

Yes. Click Open in CNAPS Studio on any card and that card's real workflow loads onto the canvas, fully wired and ready to run. It's the fastest way to learn how flows are built, and the easiest place to start customizing one.

Where do my results go?

Results from cards on Home are saved under My Work → My Creations. Flows you build on the canvas are saved separately under My Work → My Flows, and batch results under My Work → Batch Runs. You can also download any result directly from the card or the flow.

Can I switch to a light interface?

Yes. At the bottom of the left sidebar, choose Light, Dark, or System. System follows your operating system setting automatically.

Where are my workspace and billing settings?

Click your avatar in the top-right corner. Workspace Settings has four tabs — General, Billing, Usage, and Integrations — and Billing and Usage are also linked directly in that menu. Account Settings, Legal, Docs, and Sign Out are in the same place.

Image Processing Automation

Can I process many images at once?

Yes. Batch Run processes up to 100 files through the same flow in one job; for larger sets, run several batches. It supports multi-input flows with Batch / Shared / Flow Value modes per input node, plus CSV export/import for pairing inputs.

Can I chain several steps into one flow?

Yes — that's what a flow is. For example: background removal → upscaling → resize, all in a single run. To send an image down a path only when a condition is met, use Image Gate by Text: it passes the image on only when a text result — a classification label, for example — matches what you specify.

Can I replace product photo backgrounds with white or swap them for different scenes?

Yes. Remove the background and apply a clean white backdrop, or composite onto seasonal and campaign-specific scenes. Generate multiple background variations from a single product photo in one batch.

Can I use ControlNet to lock image structure while changing style?

Yes. CNAPS Studio supports end-to-end ControlNet workflows. Run any image through the Canny Edge Annotator or Depth Anything Annotator, then connect to ControlNet XL Canny or ControlNet XL Union to regenerate with a new text prompt while preserving the original structure. Use cases: sketch → photorealistic render, pose-locked character generation, style transfer with structure preservation. ControlNet XL Union accepts up to 6 control signals (canny, depth, openpose, and more) simultaneously.

Video & Audio

What video generation models are supported?

Hosted on CNAPS Studio (runs on our servers, no outside account needed):

• Helios Base: highest-quality clips, up to 60 seconds
• Helios Distilled: 2–3× faster than Helios Base, up to 60 seconds at 640×384
• Wan2.2 TI2V 5B: 5-second 720p clips from a text prompt, or animate a still image
• Cosmos3 Nano (NVIDIA): about 8-second 720p clips, with optional generated sound
• LTX-2.5: text/image/audio → video, 30s voice
• MiniMax-H3 Ref2VA: 5–15 second clips with synchronized audio, guided by an image and/or video you provide. Use it to keep a character consistent, copy motion, or lip-sync to a voice.

External (runs on Google's servers with your own Google API key):

• Google VEO 3.1: up to 4K with native audio, in 4, 6, or 8-second clips
• Gemini Omni: edit the same clip step by step ("now make it night") instead of regenerating it

Use the Video Compare tool to play up to four model outputs side-by-side, in sync, within a single flow.

Can I upscale and restore degraded video?

Yes. Supported models include SeedVR2-3B and SeedVR2-7B (single-step diffusion upscaling), FlashVSR-v1.1 (fast 4× upscaling), and SparkVSR (keyframe-guided upscaling). The video sharpening card on Home gives you one-click access without building a flow.

Can I combine image and text processing in one workflow?

Yes. Connect multimodal language models — Gemma 4, Qwen3, Claude, ChatGPT, Gemini — with image processing nodes to generate descriptions, classify content, caption images, and extract structured data, all in a single flow.

Does CNAPS Studio support document OCR and data extraction?

Yes. GLM-OCR, DeepSeek OCR, and PaddleOCR are all supported. Extract structured JSON from invoices, forms, IDs, tables, stamps, and dense document layouts.

Can I work with audio?

Plans & Pricing

What plans are available?

CNAPS Studio offers five plans: Free, Starter ($20/mo), Pro ($75/mo), Business ($200/mo), and Enterprise (custom). All paid tiers include the full model library and no watermark. For full details, see Pricing.

Is pricing flat-rate or per-image?

Flat-rate. Unlike usage-based APIs, your costs stay predictable — even during peak seasons or high-volume campaigns.

Can I try CNAPS Studio for free?

Yes, three ways:

• Without an account — run a card as a guest, three runs a day. Results are watermarked and view-only; sign up to download. Anything you made as a guest moves to your account.
• Free plan — always available, with basic capacity and every model.
• 2-week Pro trial — new users get full Pro access, including Batch Run, before deciding on a plan.

What can I do without an account?

Run cards on Home — three runs a day, with a counter on screen showing what's left. That's it, and it's enough to see what a card actually produces with your own input.

Everything else needs an account:

When you sign up, the work you did as a guest moves into your account.

Are there overage charges if I exceed usage?

No. All plans are flat-rate. Your costs stay fixed regardless of volume.

How does usage capacity work?

Capacity is measured over a rolling 5-hour period and resets automatically. Check what percentage you've used under your avatar → Usage. Running out means waiting for the reset, not paying more. Failed runs don't count against it — if a run errors out, its GPU time isn't recorded.

What's included in the Business plan?

Team workspace, SSO/SAML, webhooks, Flow Versioning, RBAC (role-based access control), and Selective Flow Sharing. $200/month.

What does the Enterprise plan offer?

On-premise deployment (including air-gapped environments) is the key differentiator — suited for healthcare, finance, defense, legal, and other regulated industries where data cannot leave the network. Pricing is custom. PoC inquiries: support@cnaps.ai.

Can multiple team members use one account?

Team collaboration is available starting from the Business plan, with SSO/SAML, RBAC, and Flow Versioning included.

Can I cancel or change my plan?

You can upgrade or downgrade at any time. Monthly subscriptions cancel at the end of the current billing cycle. Annual subscriptions cancel at the end of the contract term. Manage your plan under your avatar → Billing, or on the Pricing page.

Technical & Developers

Is there an API I can integrate with?

Yes. CNAPS Studio provides API keys that allow you to trigger Batch Run programmatically. Use the API to authenticate requests, automate workflows without opening the interface, integrate AI pipelines into your own applications, and grant third-party tools access to your flows. Generate keys under API Key in the left sidebar. For setup details, see User Guides → API Integration.

Can I use AI models without setting up CUDA, Docker, or GPU servers?

Yes. CNAPS Studio runs all models in the cloud — no environment setup needed. Skip the CUDA version conflicts, Docker configurations, and GPU server management. Just open your browser, build a pipeline, and run it.

How does the MCP integration with Claude work?

Through Claude's MCP (Model Context Protocol) integration, you can build and run CNAPS Studio flows entirely from within a Claude chat. Claude understands natural language requests, recommends the right pipeline, selects the best model for the task, diagnoses errors, and completes execution — without leaving the chat. The server exposes 34 tools, identical whether you connect via OAuth or an API key. When you connect with OAuth, the consent screen shows the redirect host and warns you if the app is unverified — check both before approving. You can see and disconnect connected apps under your avatar → Account Settings → Security → Connected apps, and disconnecting cuts that app's access immediately.

Can I use CNAPS Studio with n8n or Zapier?

Yes. The CNAPS.ai MCP Server connects to both. In n8n, an AI Agent with the MCP Client Tool node picks models and builds flows from a single chat message, authenticating with an x-api-key header. In Zapier, the MCP Client integration exposes all 34 tools as Zap actions over OAuth, so a fixed pipeline runs the same way every time. Step-by-step guides are at docs.cnaps.ai/n8n and docs.cnaps.ai/zapier.

Is on-premise deployment available?

Yes, on the Enterprise plan. For organizations where data cannot leave the network — healthcare, security, finance, defense — CNAPS Studio can be deployed on your own infrastructure in an air-gapped environment. PoC inquiries: support@cnaps.ai.

Is Edge Device deployment available?

Yes. Lightweight AI workflow pipelines can be deployed on edge devices for on-device, low-latency inference. No cloud round-trip required — suited for real-time applications, offline environments, and IoT deployments.

Can I share flows with my team?

Yes. Selective Flow Sharing supports Viewer/Editor roles, link sharing mode, access codes, and email invite notifications. You also control sharing input by input: every input node has an icon in its header that includes or excludes that node's data. Text Input is shared by default; Image Loader and Video Loader are not, so you can publish a workflow without publishing your photos or footage. Team workspaces, SSO/SAML, RBAC, and Flow Versioning are available on the Business plan.

Can I protect flows from accidental changes?

Yes. Flow Versioning (Business plan) maintains a history of flow versions and allows rollback to any prior state. RBAC lets you restrict edit permissions by role. On any flow, the History panel records every edit in real time, so you can restore a previous state with one click.

What languages does the interface support?

9 languages: English, Korean, Japanese, Chinese, Spanish, Arabic, Vietnamese, Portuguese, and French. Switch under your avatar → Account Settings → Profile → Language. Your language choice is stored on your account, not in the browser, so it follows you across every CNAPS.ai site you sign in to rather than resetting per site or per device.

What is the Open Source AI Server?

CNAPS Studio's AI server infrastructure is being released as an open-source project, enabling developers to self-host the AI execution layer, inspect the codebase, and contribute. Public repository launch is targeted soon.

Security & Data

Is my image data handled securely?

Uploaded images are not shared with anyone outside CNAPS, and we do not use them to train AI models. Two exceptions, both under your control:

• External model nodes. If your flow uses an external model (ChatGPT, Claude, Gemini, GPT Image, Nano Banana, VEO, Omni, Gemini TTS), CNAPS sends that node's input to the provider's API using the key you configured, and the provider handles it under its own privacy policy. Remove the key in Workspace Settings → Integrations to disable those nodes.
• Content screening. Prompts, uploads, and outputs are screened for prohibited content by us and by service providers acting on our behalf, which process the content solely to perform the screening.

For stricter compliance requirements, on-premise deployment is available through the Enterprise plan — data never leaves your infrastructure.

Can I auto-blur or mosaic faces across hundreds of photos?

Yes. CNAPS Studio uses AI-powered face detection to automatically blur or mosaic faces in images. Batch Run handles up to 100 photos per batch.

Is there a solution for comprehensive privacy protection in images?

Yes. BlindSpot is a standalone web app powered by CNAPS Studio AI models. Features include face anonymization, tattoo concealment, license plate blur, text redaction, and body silhouette. Suited for media companies, legal teams, and compliance workflows.

Can I find where a logo or a person appears in a video?

Yes, in two ways:

• SAM 3.1 (Video) tracks and masks any object you describe in text — "red car", "person in a blue jacket" — across every frame of a clip.
• Video Analysis (Gemini) answers questions about a video in plain language and returns timecodes for the moments you ask about. It's an external model, so it runs on your own Google API key.

Why was my request rejected?

CNAPS Studio screens prompts, uploaded files, and generated outputs against the prohibited-content rules in our Terms of Service. Screening is automated, and runs both before a flow starts and before an output is stored, so a request can be refused at either point. We may refuse to process, generate, store, or deliver content that screening identifies as prohibited, without prior notice.

A refused request is never charged. It consumes no capacity, whether the refusal happens before or after generation — including when screening was temporarily unavailable.

If you think a request was refused in error, email support@cnaps.ai. We will review the refusal and respond.

Can I ask a person to look at it, rather than the system?

Yes. Screening is automated, and you have the right to contest an automated refusal and ask for human review. Email support@cnaps.ai with the flow and the inputs you used.

What content isn't allowed?

The rules are in Section 4 (Prohibited Activities) of our Terms of Service, and the additional rules for anything you publish to Community are in Section 8.8. Our AI Policy is incorporated into the Terms and applies as well; where two terms conflict, the more restrictive one applies.

Some models carry their own acceptable-use terms on top of ours — they are listed on that model's page in the AI Model Hub, and using the model means agreeing to them.

Can content I already shared be removed?

Yes. We may remove or disable access to content that violates the Terms, and may issue warnings or suspend or terminate accounts.

One thing to know about Community: if a Workflow you published is removed or unpublished, copies other users forked before the removal stay in their workspaces. Those are independent copies, outside our control. Removal stops new forks and takes the Workflow out of public display; it does not reach back into copies already made.

What happens to my account if I break the rules?

Depending on what happened, we may remove the content, issue a warning, or limit, suspend, or terminate your access to the Services.

For apparent child sexual abuse material, we report to the National Center for Missing & Exploited Children (NCMEC) and to law enforcement as required by law, preserve the related content and account information for that purpose, and terminate the associated accounts.

Is my content stored when it's refused?

We keep records of screening decisions. Content and account information associated with suspected CSAM is retained for as long as the law requires, including the preservation period that applies to NCMEC reports.

Screening is carried out by us and by service providers acting on our behalf, which process the content solely to perform the screening. See our Privacy Policy, Sections 4, 5.3, and 9.

AI-Generated Content Labels & Transparency

Why is there an "AI-generated" badge on my output? Can I turn it off?

The badge is applied automatically to satisfy EU AI Act Article 50, which requires AI-generated content to be identifiable. It doesn't disappear after download — it persists in three places: a visible badge on the Output Viewer node in the canvas, provenance metadata embedded in the file itself when you download it, and a field in the API response when you fetch it programmatically. None of the three can be turned off — it's a compliance requirement, not a display option.

Why does my segmentation or edited image still show a generative model's name in the badge?

CNAPS Studio tracks provenance across the entire flow, not just the last node. If FLUX generated the image and BiRefNet then segmented it, the final output is a segmentation of an AI-generated image — which still counts as AI-generated content. The badge attributes to the earliest AI generation step in the upstream chain.

My output has no badge. Is something wrong?

No. Your flow doesn't include any generative or edit-type AI model — for example, a flow that only upscales or denoises an original photo. No new content is generated, so no badge appears.

How do I verify the provenance metadata in a downloaded file?

Use exiftool <filename> for images, videos, or audio files to inspect EXIF, MP4 atom, or ID3 fields. For .txt files, open the file in any text editor and look at the footer at the end. For API responses, check the X-CNAPS-Generated-By HTTP header.

A customer or auditor asked for our AI transparency documentation. What should I send them?

Direct them to our AI Policy page — it covers what AI models we use, how outputs are marked, our risk classification approach, and acceptable use restrictions. For deeper compliance questions or enterprise-specific documentation, contact legal@cnaps.ai.

Getting Started & Support

How do I get started?

Just open cnaps.ai — you land on Home and can run a card as a guest, three runs a day, no account needed. Pick a card, look at the finished example, replace the input with your own, and click Run again — your first result takes about 30 seconds. Sign up when you want to download what you made, open the canvas, or keep a library — what you made as a guest comes with you. Step-by-step instructions are in the Quickstart Guide.

Where can I find the documentation?

Click Guides at the bottom of the left sidebar, or your avatar → Docs for direct links to Guides, Quickstart, Models, and FAQ. Everything is also at docs.cnaps.ai.

How do I get help?

Email support@cnaps.ai for product questions and PoC inquiries, or legal@cnaps.ai for compliance and licensing questions.

CNAPS Studio vs. Other Tools

How is CNAPS Studio different from n8n?

n8n is a general-purpose workflow automation platform — great for connecting apps, APIs, and business logic. However, it's not built for vision AI. n8n lacks native support for AI models like object detection, segmentation, or image enhancement, and its image handling is limited to basic operations. CNAPS Studio is purpose-built for image and video processing: pre-integrated vision AI models, a visual pipeline builder designed for image workflows, native Batch Runs for up to 100 files per batch, and no need to wire up external APIs for core processing tasks.

The two also work together: an n8n AI Agent can drive CNAPS Studio through the MCP Server. See Connect CNAPS.ai to n8n.

How is CNAPS Studio different from ComfyUI?

ComfyUI is a node-based tool for AI image generation using Stable Diffusion. CNAPS Studio covers a broader scope: image processing and analysis (background removal, object detection, segmentation, OCR, upscaling) plus video generation and upscaling. Running ComfyUI yourself means managing a GPU and CUDA environment. CNAPS Studio runs entirely in the cloud. CNAPS Studio also has Batch Run built in; ComfyUI batch workflows often require custom nodes and manual configuration.

How is CNAPS Studio different from RunwayML / Pika?

RunwayML and Pika are single-model video generation UIs. CNAPS Studio is a workflow-first platform: build custom pipelines combining multiple models, process batches at scale, and control outputs deterministically. Beyond video generation, CNAPS Studio chains video upscaling, frame extraction, VLM analysis, and OCR in a single flow.

How is CNAPS Studio different from MindStudio / BuildShip / Dify?

MindStudio, BuildShip, and Dify are general-purpose AI agent automation or app builders. CNAPS Studio specializes in image and video processing with 70+ pre-integrated vision AI models — object detection, segmentation, OCR, video upscaling, and more — all built in, with no external API wiring needed for core tasks.

How is CNAPS Studio different from Zapier?

Zapier automates data movement between apps. CNAPS Studio runs AI models on images and videos in batch. They serve different purposes: use CNAPS Studio for image and video AI processing, use Zapier for app-to-app workflow automation. They're also complementary: the CNAPS.ai MCP Server connects to Zapier, so a Zap can run any CNAPS Studio flow. See Connect CNAPS.ai to Zapier.

How is CNAPS Studio different from Adobe Firefly / Photoshop AI?

Adobe Firefly and Photoshop AI are editing tools built around a single image at a time, driven by a human at the keyboard. CNAPS Studio is built around repeatable pipelines: define the sequence once, then run it across whole folders of files unattended. It also gives you model choice — open-source and external models side by side, swappable per node — rather than one vendor's model. For a designer refining one hero image, Photoshop is the better tool. For a team processing an entire catalog the same way every week, CNAPS Studio is.

Last Updated: Oct 16th, 2026