Digitize objects, plants and scenes from multi-view photos or video on Windows. Build point clouds with COLMAP photogrammetry or reconstruct with NeRF and 3D Gaussian Splatting. Preview and export results.
Compute
NVIDIA GPU Required
Video memory
12GB+ VRAM
Image and model size
Image >10GB
Cutout & Watermark Removal
Released
Install and launch Cutout & Watermark Removal from Novaber, the local AI workspace for Windows. Batch remove photo backgrounds or clear visible watermarks from images and short clips. Export transparent PNGs, masks, or restored media.
Compute
NVIDIA GPU required
Video memory
8 GB+ VRAM recommended
Image and model size
Image downloads ~16 GB
OCR Document Recognition
Released
Run batch OCR on scanned PDFs and images locally. Extract text and tables with PaddleOCR, GLM-OCR or OvisOCR2, review against the source, and export Word, Excel or searchable PDF files.
Compute
NVIDIA GPU required (CUDA 12.6)
Video memory
VRAM varies by model and page complexity
Image and model size
Backend/worker ~18.1 GB unpacked each; models extra
Voice Studio
Released
Turn text, TXT scripts or SRT subtitles into video narration, course audio and multi-character dialogue with local Qwen3-TTS. Correct pronunciation, redo individual lines, and export WAV, MP3, subtitles or separate role tracks.
Compute
NVIDIA GPU recommended
Video memory
4 GB standard voices; 8 GB recommended for clone/design
Image and model size
Shared image layers use about 7 GB; the three models require about 9.6 GB more
AI Image Studio
In development
In development for creating covers, event posters and marketing graphics. Planned features include text-to-image with FLUX.2 klein and Z-Image-Turbo, reference-image editing with Qwen-Image-Edit, editable Chinese titles, and landscape and portrait exports.
Compute
NVIDIA GPU required
Video memory
About 13GB minimum for FLUX.2 klein; 24GB recommended for Qwen editing
Image and model size
Application images use about 4.5GB; the full three-model cache is about 99.3GiB
Privacy Redaction
In development
In development for preparing images and short videos for sharing. Planned features include batch face blurring, face and license plate pixelation, manual regions, review by timestamp, and local exports.
Compute
CPU only
Video memory
No dedicated VRAM required
Image and model size
Three application images use about 485 MB unpacked; the on-demand model cache uses about 4.2 MB
Video Localization
In development
In development for creators publishing Chinese videos to international audiences. Planned features include translating Chinese subtitles into English, reviewing subtitles alongside the video, and exporting English or bilingual subtitles or a captioned video.
Compute
CPU supported; no discrete GPU required
Video memory
At least 2 GiB of available system memory recommended
Image and model size
Models use about 1.3 GiB; image size pending build verification
Collected Memories
In development
Import HEIC originals, find photos by describing the scene, and select travel or family photos on a timeline. Edit album chapters and captions, then export a printable PDF, an offline album or a package of originals.
Compute
NVIDIA GPU Required
Video memory
12GB+ VRAM recommended
Image and model size
Media image about 14GB unpacked; photo / video / image-text models add about 6.2GB / 6.1GB / 1.5GB
Before you get started
Where do these AI applications run?
Applications are deployed and run locally through Novaber on Windows, with Docker providing separate application environments. Initial downloads, sign-in and some features need a network connection; check each application for details.
How can I check whether an application suits my computer?
Compare the GPU, VRAM and storage requirements shown on each card with the detailed requirements. NVIDIA GPU and VRAM needs vary by application, so prepare the environment for the tool you choose.
Can I install applications that are still in development?
Applications marked as in development are not available to use yet. Choose a released application and follow its setup guide to prepare Novaber and the required environment.