supportsage 0.6.3

AI-optimized support structures for 3D printing

SupportSage 🧠🖨️

AI-optimized support structures for 3D printing.

Stop guessing. Stop wasting filament. Stop sanding support scars.

SupportSage analyzes your STL model and generates intelligent, geometry-aware support structures that use 30-50% less material, reduce print time, and leave smoother surfaces.

🚀 Install

pip install supportsage

Or from source:

git clone https://github.com/bossman-lab/supportsage
cd supportsage && pip install -e .

---

Features

Feature	CLI	Web Demo	GUI	Cura Plugin	Prusa Plugin	Orca Config
Overhang analysis	✅	✅	✅	✅	✅	✅
Tree support generation	✅	✅	✅	via CLI	✅	✅
Material savings estimate	✅	✅	✅	✅	✅	✅
Interactive 3D viewer	❌	✅	✅	❌	❌	❌
One-click optimize	❌	❌	✅	✅	✅	✅
Batch processing	✅	❌	❌	❌	❌	❌
JSON export	✅	❌	✅	✅	✅	✅
FilamentDB integration	✅	✅	✅	❌	❌	❌
Print inspection (Printsight)	✅	✅	✅	❌	❌	❌
Multi-material interface	✅	❌	❌	❌	❌	❌
GCode preflight	✅	❌	❌	❌	❌	❌
Printability score	✅	❌	❌	❌	❌	❌

Quick Start

# Analyze overhangs
supportsage analyze model.stl

# Generate optimized tree supports
supportsage tree model.stl -o optimized.stl --strategy balanced
#                      strategy: light | balanced | heavy

# Generate pillar supports
supportsage optimize model.stl -o model_with_supports.stl --strategy balanced

# Export strategy for Cura plugin
supportsage export model.stl -f json -o strategy.json

Integration Suite 🔗

SupportSage integrates with two companion tools for a complete 3D printing workflow:

FilamentDB — Smart Parameter Lookup

Pass --filament "Brand Model" to any command. SupportSage auto-looks up the recommended nozzle temperature, bed temperature, and other settings from the FilamentDB open database.

# Analyze with filament-specific recommendations
supportsage analyze model.stl --filament "Bambu Lab PLA Basic"
# → Adds: 🔥 Nozzle: 220°C | Bed: 55°C | Fan: 100%

# Optimize with filament context
supportsage tree model.stl -o optimized.stl --filament "eSun PLA+"

Printsight — Post-Print Quality Check

Pass --inspect photo.jpg to the tree command. After generating supports, it runs a full Printsight quality inspection on your printed result.

# Generate supports → then inspect the print
supportsage tree model.stl -o optimized.stl --filament "PLA+" --inspect after_print.jpg
# → 🌳 Tree supports generated
# → 🔥 Filament: 215°C / 60°C
# → 🖨️ Printsight: Good (0.82) — No stringing, minor warping

Complete Workflow — One Command

# Analyze → Print → Inspect — all in one pipeline
supportsage tree model.stl -o out.stl --filament "Bambu Lab PLA Basic" --inspect final_print.jpg

Desktop GUI

supportsage-gui --open
# → Opens http://localhost:5000

A local web app that combines the 3D viewer with the full CLI engine. Drop STL → visualize overhangs → click "Generate Tree Supports" → download optimized STL. No terminal needed beyond the launch command.

Cura Plugin

Adds "SupportSage" menu to Cura:

• Run SupportSage on Model... — analyzes the open STL and applies settings automatically
• Load Strategy... — imports a pre-exported JSON strategy
• Clear SupportSage Settings — resets to defaults

Install: bash cura_plugin/install.sh or copy cura_plugin/ to Cura's plugins folder.

Benchmarks

Model	Traditional	SupportSage	Savings
Multi-bridge	6,317mm³	4,211mm³	33%
Cantilever platform	18,440mm³	12,293mm³	33%
Multi-level scaffold	11,194mm³	7,463mm³	33%
Total	35,951mm³	23,967mm³	33%

Architecture

STL input → trimesh mesh
     ↓
Overhang detection (face normal × Z-up dot product)
     ↓
Severity classification (critical / moderate / borderline / safe)
     ↓
Island detection (disconnected region BFS)
     ↓
Per-island strategy assignment
     ↓
Tree support generation (branching organic structures)
     ↓
STL / JSON / G-code export

How It Works

Traditional slicers use a binary check: "Is this face >45° from vertical? Fill everything beneath with support."

SupportSage treats overhangs as a spectrum:

Severity	Angle	Strategy
🔴 Critical	>120° from vertical	Dense interface, thick branches
🟠 Moderate	70-120°	Tree supports, medium density
🟡 Borderline	45-70°	Light touch, thin pillars
🟢 Safe	<45°	No support needed

Then it finds disconnected overhang regions (islands) via BFS and assigns the optimal strategy per island. A bridge span gets two small support islands at each end — not one giant block.

Web Demo

bossman-lab.github.io/supportsage (enable GitHub Pages first)

A pure client-side version: drag an STL, see color-coded overhangs, rotate/zoom, no upload required. Works entirely in your browser via Three.js.

CLI Reference

supportsage analyze model.stl          Analyze overhangs
supportsage analyze model.stl --filament "PLA"   + filament recommendations
supportsage tree model.stl -o out.stl  Generate tree supports
supportsage tree --dual-material model.stl  + soluble interface layer (v0.3)
supportsage preflight model.gcode      Pre-print GCode risk analysis (v0.5)
supportsage precheck model.stl         Pre-slicing printability check (v0.5)
supportsage tree model.stl --filament "..." --inspect photo.jpg   full pipeline
supportsage optimize model.stl -o...   Generate pillar supports
supportsage export model.stl -f...     Export support strategy
supportsage-gui --open                 Launch desktop GUI (with filament + inspect)

Related Projects

• FilamentDB — Open-source filament parameter database. Search, compare, get AI-recommended print settings.
• Printsight — 3D print quality inspection. Detect stringing, layer issues, warping from a photo.

Blog Posts

1. I Built an AI That Slices Your Support Waste by 40%
2. How Tree Support Generation Actually Works
3. We Benchmarked SupportSage: Here's the Data
4. Browser-Based 3D Print Overhang Analyzer
5. Desktop GUI and Cura Plugin

License

MIT