ImagingCPU3 credits

Cellpose-SAM — Cell & Nucleus Segmentation

Segment cells and nuclei in any microscopy image

Cellpose-SAM combines the Cellpose generalist segmentation framework with Meta's Segment Anything foundation. Upload any microscopy image, get instance masks, bounding boxes, areas, and per-object mean intensity.

$0.03
per API call
3
credits per call
/v1/imaging/cellpose
API endpoint

Features

Base64 image upload (PNG/JPEG)
5 model variants: cyto3, cyto2, nuclei, livecell, tissuenet
Per-object bounding boxes + areas + mean intensity
Configurable diameter parameter
CPU-only — no CUDA setup needed

Quick Start

Cellpose-SAM — Python Examplepython
import requests, base64

with open("microscopy.png", "rb") as f:
    image_b64 = base64.b64encode(f.read()).decode()

result = requests.post(
    "https://scirouter.ai/v1/imaging/cellpose",
    json={"image_b64": image_b64, "diameter": 30, "model": "cyto3"},
    headers={"Authorization": "Bearer sk-sci-your-key"}
).json()["data"]
print(f"Detected {result['n_objects']} objects")

Use Cases

1

Cell counting in histology or IHC

2

Nuclear segmentation of DAPI images

3

Live-cell time-lapse analysis

Start Using Cellpose-SAM

500 free credits every month. No credit card required.