Changelog
Get the latest updates and improvements to our platform.
Solar Scout: AI-Powered Rooftop Analysis Tool
Introducing Solar Scout! Use AI technology to quickly analyze rooftop areas in satellite imagery, helping you evaluate solar panel installation potential.
Key features include:
- Smart Rooftop Detection - Automatically detect and mark rooftop areas in satellite images
- Batch Scanning - Analyze multiple locations simultaneously
- History Management - Save and review all your scan results
- Map Search - Search by address to quickly locate target areas
- Multi-language Support - Available in English, Chinese, Japanese, Korean, and more
Official Node.js Demo & API Documentation Enhancement
We've released an official Node.js demo application to help developers quickly integrate the SAM 3 API. The demo includes complete Express.js backend and interactive frontend implementations.
Key updates include:
- Official Node.js Demo - Complete working example with backend and frontend
- API Documentation Enhancement - Added code examples and troubleshooting guides
- Quick Start Guide - Step-by-step integration instructions
- Image Proxy Optimization - Improved image loading with thumbnail support
LINE Sticker Maker: AI-Powered Sticker Generation
Introducing the LINE Sticker Maker! Generate custom LINE stickers using AI with reference images and text prompts. Features automatic background removal and sticker cutting powered by SAM 3.
Key features include:
- AI Sticker Generation - Create stickers with text prompts and reference images
- Smart Segmentation - Automatic object extraction with SAM 3 technology
- LINE Format Export - Stickers optimized for LINE's size requirements
- History Management - Save and manage your sticker generation history
- Batch Download - Download all stickers as a ZIP file
SAM 3 API Enhancements & Polygon Merging
We've enhanced the SAM 3 API with raw and workflow routes and added support for merging polygon objects. This update makes processing complex object segmentation more flexible and efficient.
Key updates include:
- Added raw and workflow API routes
- Supported polygon object merging function - Merge multiple split objects into one
- Refactored prediction processing logic - Improved efficiency and accuracy
- Optimized API image size handling - Dynamic resizing to save bandwidth and credits
SegmentLens Pro Online: Text-Prompt Smart Segmentation Service Launch
We're excited to announce the official launch of SegmentLens Pro Online Service! This cloud-based, high-performance image segmentation solution supports text-prompt smart segmentation, automatic object detection, history management, and more.
Key features include:
Sign up to get free credits, then continue with credits.
- Text-Prompt Segmentation - Enter English descriptions to automatically identify and extract objects
- Smart Object Detection - Automatically recognize all objects in images and fill prompts
- History Management - Browse, restore, and delete segmentation history with pagination
- Mobile Support - Perfectly adapted for mobile devices and desktops
- Batch Download - Download all segmented objects with one click
- User Onboarding - Interactive tour to help you get started quickly
SegmentLens Playground: Free In-Browser Image Segmentation
We've launched the SegmentLens Playground! You can now try SegmentLens image segmentation directly in your browser. It runs locally using WebGPU, ensuring your privacy as no images are uploaded to our servers.
Features include point and box segmentation, multiple model size options (q4 to fp32), and instant mask downloads.
Introducing SegmentLens with SAM 3 Technology
SAM 3 (Segment Anything Model 3) brings text prompts and exemplar-based segmentation to image understanding. The 848M parameter model features text prompts for natural language object description, exemplar prompts for similar object detection, visual prompts with point and box selection, and video tracking across frames.
SegmentLens combines SAM 3 model segmentation with Gemini generation for e-commerce product background removal and education content creation. API access is now available for developers.