SegmentLens

SegmentLens is an AI-powered image segmentation platform that uses Meta's SAM 3 model to detect, mask, and edit objects in images and videos with pixel-level precision.

What is SegmentLens?

SegmentLens is a web-based and iOS/macOS app that performs image segmentation using the SAM 3 model combined with Google Gemini for generation tasks. It takes images (uploaded or from examples) and user prompts (text, clicks, or boxes) as input and outputs segmentation masks and edited images. It is developed by an independent team and is not affiliated with Meta or Google.

Key Features

• Pixel-level segmentation — Detects objects, backgrounds, and text regions in seconds with over 98% accuracy.
• Multi-model generation — Uses Google Gemini alongside SAM 3 for background replacement, style transfer, and creative generation from a single image upload.
• Text prompts — Mask all objects matching a short English phrase, like “dog” or “car.”
• Visual prompts — Segment objects using positive/negative clicks or by drawing a bounding box around an example.
• Video tracking — Track segmented objects across video frames with stable temporal tracking (SAM 3 capability).
• API integration — Standardized HTTP API that returns segmentation masks; integrate with three lines of code.
• Interactive refinement — Add follow-up prompts to correct or refine segmentation results.
• Free trial credits — New users receive free credits to test the platform without commitment.

Who is it for?

• E-commerce professionals — Remove backgrounds or swap scenes for product photos at scale.
• Graphic designers — Generate precise masks for compositing and creative edits.
• Developers — Integrate segmentation into apps or websites via the API.
• Content creators — Segment objects in videos for editing or effects.

What can you do with SegmentLens?

• Remove backgrounds — Instantly isolate subjects from any image with a single click.
• Generate new backgrounds — Replace the background with a different scene using Gemini generation.
• Segment multiple objects — Use text or exemplar prompts to detect and mask all instances of a category.
• Edit with follow-up prompts — Refine masks interactively without restarting.

How does SegmentLens work?

1. Upload an image to the web app or use the iOS/macOS app.
2. The SAM 3 model analyzes the image and generates segmentation masks based on your prompt (text, clicks, or box).
3. For generation tasks, Gemini processes the masked areas to produce new backgrounds or styled outputs.
4. For API users, send an HTTP request with the image and prompt; receive a segmentation mask in seconds.

Pricing

SegmentLens offers three monthly credit tiers: Basic (300 credits/month), Pro (600 credits/month), and Max (1500 credits/month). Yearly plans are available at a 20% discount. Exact dollar prices are not displayed on the site.

Pros and cons

Pros: High segmentation accuracy (98%+), supports text, visual, and exemplar prompts, video tracking, and offers an API for integration. Cons: Only English prompts are supported; full segmentation requires signing in; free credits are limited per tier.