SentiSight.ai
sentisight.ai
- Free plan
- Linux
- Web
- iPhone and iPad
- Android
- API
- Self-hosted
What SentiSight.ai is
SentiSight.ai is a web-based platform for labeling images and building AI image-recognition applications. It supports classification, object detection, and segmentation, with annotation tools for labels, boxes, polygons, points, polylines, and bitmaps. Smart labeling uses selected foreground and background points to extract an object from an image. Teams can work in shared projects, where managers track progress and time and manage roles and permissions. Image search can find results resembling an uploaded query or similar pairs within a dataset. Pre-trained models cover content moderation, goods classification, automatic hashtags, and people counting. Models can run through the web interface or REST API, or offline after downloading a model and setting up a local REST API server. Android and iOS apps make predictions from uploaded images or camera photos. Those apps cannot train models or label images already uploaded, and support classification labels but not object labels. The platform has a free plan, a 30-day trial, and paid pricing from €0.10/mo.
Who it suits
It suits teams labeling image datasets or developing recognition models for uses such as retail tagging, manufacturing quality checks, visual search, and research. Mobile users can make predictions, but training and labeling uploaded images require another platform option.
Good to know, and what to watch
- Supports classification, detection, and segmentation models
- Offers multiple annotation types and smart labeling
- Shared projects include roles and progress tracking
- Models can run online or offline
- Pre-trained models cover several image tasks
- Mobile apps cannot train new models
- Mobile apps cannot label uploaded images
- Mobile apps support classification, not object labels
Plans
Every plan and priceOffline model license
Pay-as-you-go
Free plan · paid from €0.10/mo · 30-day trial
Getting started
- Open the SentiSight.ai website and create an account
- Start with the free plan or the 30-day trial
- Create a project and upload images for labeling
- Choose annotation tools or smart labeling for objects
- Train or select a model and deploy through the web interface or REST API
- For offline use, download a model and configure a local REST API server
Limits to check
Pay-as-you-go includes €5 monthly complimentary credits and €20 free credits for new users, covering 5,000 predictions, 5,000 labels, or 83 minutes of training within those credits. Additional listed rates are €1 per 1,000 predictions, €1 per 1,000 labels, and €3.6 per training hour, with 5 GB disk space included. Mobile apps cannot train models or label already uploaded images and do not support object labels.
The Club's review
Overview
SentiSight.ai is an image-recognition platform for teams that need to prepare datasets, train custom models, and put image analysis into production. It is best suited to developers, researchers, and operational teams whose workflow spans annotation and deployment rather than to users seeking a phone-only tool. Its broad model and deployment coverage is a strong reason to consider it, while metered usage and limited mobile authoring may rule it out for some buyers.
Its stated applications include retail product tagging, manufacturing quality inspection and defect detection, e-commerce visual search, and research dataset labeling. The platform is associated with Neurotechnology UAB, founded in 1990 and headquartered in Vilnius, Lithuania. For more options, see AI Image Recognition Software.
Key features
SentiSight.ai supports image classification, object detection, and segmentation. Its annotation workspace handles classification labels, bounding boxes, polygons, points, polylines, and bitmaps, so teams can choose markings appropriate to simple categories, object outlines, or pixel-level work. Smart labeling uses user-selected foreground and background points to extract an object; this can reduce repetitive tracing, although it still requires a person to guide the selection.
Shared projects support multi-user labeling. Project managers can monitor progress and time and administer roles and permissions, making the service a better fit for supervised dataset production than an isolated personal project. Similarity search finds images resembling an uploaded query image, while NvN searches retrieve similar image pairs within a dataset. Those tools are useful for visual discovery, duplicate-like matching, and product-search workflows.
Pre-trained models cover tasks including content moderation, goods classification, automatic hashtags, and people counting. They can shorten the path to a useful result when a ready-made task matches your needs, while custom models cover projects that require your own data. Models can run online in the web interface or through a REST API, or offline after downloading a model and configuring a local REST API server. That hybrid approach suits organizations that need local execution as well as hosted access.
Android and iOS apps can make predictions from uploaded images or camera photos. They cannot train new models or label images that have already been uploaded, and they support classification labels but not object labels. Mobile is therefore useful for prediction checks and field capture, not as a replacement for the desktop labeling and training workflow.
Pricing
SentiSight.ai uses a freemium model. A free plan is available, paid access starts from €0.10/mo, and a 30-day trial is offered. The published entry price alone does not describe a complete production allowance, so teams should match expected usage to the metered terms below.
Pay-as-you-go
This plan includes €5 in monthly complimentary credits and €20 in free credits for new users. The free credits cover 5,000 predictions, 5,000 labels, or 83 minutes of training. Beyond that allowance, predictions cost €1 per 1,000, image labeling costs €1 per 1,000 labels, and training costs €3.6 per hour. The plan includes 5 GB of disk space. It is the clearest fit for variable workloads and experimentation; costs can rise with large datasets, repeated training, or high-volume inference.
Offline model license
This plan is billed as a single time payment, with a custom quote based on model speed. It includes offline model download and a 30-day free trial. Choose it when local REST API deployment is a requirement and a one-time license is preferable to ongoing hosted usage charges. The custom quote means budgeting requires a conversation with the provider.
The free allowance is useful for evaluating the workflow, but teams should not assume it will cover sustained labeling or training.
Platforms
SentiSight.ai is available on the web and supports Android, iOS, Linux, a REST API, and self-hosted deployment. This range covers browser-based project work, mobile prediction, programmatic integration, and local serving. The mobile restrictions make web, API, or self-hosted access the practical choice for teams doing serious annotation and model development.
Who it's for
Choose SentiSight.ai for a retail, manufacturing, e-commerce, or research workflow that needs both structured labeling and model delivery. It is particularly well matched to teams coordinating annotators, tracking project work, and requiring online and offline deployment paths. It is less suitable when all labeling and training must happen on mobile devices, or when predictable flat-rate pricing matters more than flexible usage billing.
Pros and cons
Pros
- Broad annotation coverage: boxes, polygons, points, polylines, bitmaps, and classification labels support varied computer-vision datasets.
- Team controls: shared projects, progress and time tracking, and role permissions support managed labeling operations.
- Flexible deployment: web, REST API, and downloadable offline models accommodate hosted and local environments.
- Useful discovery tools: image similarity and NvN search support visual-search and dataset-investigation work.
- Low-risk entry: a free plan, introductory credits, and a 30-day trial allow evaluation before larger spend.
Cons
- Mobile authoring is restricted: the apps cannot train models or label uploaded images and lack object labels.
- Usage is metered: training, labeling, and predictions incur separate charges after credits, complicating costs for intensive projects.
- Offline pricing is custom: buyers needing local deployment cannot calculate the license cost from a published rate.
- Free capacity is finite: 5,000 predictions, 5,000 labels, or 83 training minutes may be quickly consumed by substantial datasets.
Alternatives
Sighthound AI is a better fit when you need a paid platform with Android, iOS, Windows, macOS, Linux, API, web, and self-hosted coverage, including a free tier limited to one user and two minutes of access.
Face++ suits buyers seeking a freemium API with a free key and API-group-specific QPS limits.
Google Cloud Vision API is preferable when image features are consumed as pay-as-you-go API units, with the first 1,000 units free.
Yandex Vision OCR is the focused alternative for paid OCR requests such as printed text or license plates.
CompreFace is the choice for free, open-source, Docker-based deployment with CPU and GPU support.
InsightFace fits research and non-commercial projects seeking an open-source option, while commercial model use requires a separate license.
MediaPipe is suited to free on-device machine-learning solutions and framework components.
Roboflow is worth choosing when its free tier of 10 monthly credits or its $39-per-month Core plan matches your training and inference workflow.
Verdict
SentiSight.ai is a strong choice for teams that want one platform for detailed image annotation, custom model work, similarity search, and both hosted and offline deployment. Its main reason to choose it is workflow breadth backed by collaboration and hybrid serving; its main reason to look elsewhere is the combination of mobile restrictions, finite free allowances, and usage-based costs. Start with the free plan or 30-day trial, then move to pay-as-you-go for variable demand or request an offline license when local serving is essential.
The verdict
SentiSight.ai combines dataset labeling, model development, image search, and deployment options. Check the mobile limitations if labeling or training from a phone is important.
Questions people ask
Does SentiSight.ai have a free plan or trial?
Yes. It offers a free plan and a 30-day free trial for the offline model license.
What are the pay-as-you-go allowances and rates?
New users receive €20 in free credits, and the plan includes €5 monthly complimentary credits. Credits cover 5,000 predictions, 5,000 labels, or 83 training minutes; listed rates afterward are €1 per 1,000 predictions, €1 per 1,000 labels, and €3.6 per training hour.
Which platforms are supported?
SentiSight.ai supports Android, iOS, Linux, web, API, and self-hosted deployment.
Can models run offline?
Yes. Download a model and set up a local REST API server for offline operation. The offline model license includes a 30-day free trial, with a custom quote based on model speed.
What can the mobile apps do?
Android and iOS apps make predictions from uploaded images or device-camera photos. They cannot train new models or label existing uploads, and they support classification labels rather than object labels.
What the maker publishes
- Annotation
- Its labeling tool supports classification labels, bounding boxes, polygons, points, polylines, and bitmaps. source
- Collaboration
- Shared projects let multiple users label images, while project managers can track progress and time and manage roles and permissions. source
- Data processing
- The data transfer agreement identifies Neurotechnology UAB as a GDPR data processor and says it will process personal data on the user's instructions and bind authorized personnel to confidentiality. source
- Deployment
- Models can run online through the web interface or REST API, or offline after downloading a model and setting up a local REST API server. source
- Founded
- 1990 source
- Headquarters
- Vilnius, Lithuania source
- Intended use
- The maker describes use cases including retail product tagging, manufacturing quality control and defect detection, e-commerce visual search, and research dataset labeling. source
- Mobile apps
- The mobile app is available for Android and iOS and supports predictions using uploaded images or photos taken with the device camera. source
- Mobile limitations
- The app cannot train new models or label already uploaded images, and it supports classification labels but not object labels. source
- Model types
- The platform supports classification, object detection, and segmentation models. source
- Pre-trained models
- Pre-trained models are available for tasks including content moderation, goods classification, automatic hashtags, and people counting. source
- Privacy
- The privacy policy says personal-data safeguards are handled in accordance with GDPR and other applicable data protection laws. source
- Purpose
- SentiSight.ai is a web-based platform for image labeling and developing AI-based image recognition applications. source
- Similarity search
- Users can search for images similar to an uploaded query image and run NvN searches to retrieve similar image pairs in a dataset. source
- Smart labeling
- The smart labeling tool uses selected foreground and background points to let AI extract a labeled object from an image. source
- Support
- The site provides a FAQ, user guides, video tutorials, and a contact page for assistance. source
Checked against the maker's pages on 30 Sept 2026.