Artifactiq Releases

Public releases for Artifactiq Core - the high-performance AI-powered visual intelligence engine.

Prerequisites

mlOS Axon v3.3.0+ is required for model management with YOLO ONNX support. Install it first:

curl -sSL axon.mlosfoundation.org | sh

Learn more at mlosfoundation.org

Quick Start

Step 1: Install Axon (Required)

curl -sSL axon.mlosfoundation.org | sh

Step 2: Install Artifactiq

curl -fsSL https://artifactiq.ai/install.sh | sh

This automatically detects your platform and installs to ~/.local/bin.

Step 3: Download a Model

# Install YOLOv8 nano (fastest, 6MB) with ONNX conversion
axon install hf/ultralytics/yolov8n --format onnx

Step 4: Analyze an Image

artifactiq analyze --input photo.jpg --model yolov8n

Validate Your Installation

After installation, run these commands to verify everything is working:

# Check Artifactiq version
artifactiq --version
# Expected: artifactiq 1.0.0-alpha.13

# Check model backend status
artifactiq models
# Should show Axon is configured

# List installed models
axon list
# Should show yolov8n if you installed it

# Test with a sample image (download a test image first)
curl -o test.jpg https://ultralytics.com/images/bus.jpg
artifactiq analyze --input test.jpg --model yolov8n
# Should detect: bus, person, etc.

Expected Output

Analysis Results
================

Detected 4 objects:
  - person (95.2%)
  - person (91.7%)
  - bus (89.3%)
  - person (85.1%)

Processing time: 47ms

Batch Processing (NEW in v1.0.0-alpha.13)

Process entire directories of images with a single command:

# Analyze all images in a directory
artifactiq analyze --input ./my-images/

# JSON output for batch results
artifactiq analyze --input ./my-images/ --format json

Batch Output

=== photo1.jpg ===
Detected 3 objects:
  - person (92.1%)
  - car (87.5%)
  - dog (76.3%)
Processing time: 42ms

=== photo2.webp ===
Detected 1 objects:
  - person (89.2%)
Processing time: 38ms

================
Batch Summary
================
  Images processed: 2/2
  Total time: 80ms
  Avg per image: 40ms

Supported Formats

jpg, jpeg, png, webp, bmp, gif, tiff, tif

CoreML Support for Apple Silicon (NEW in v1.0.0-alpha.13)

On Apple Silicon Macs, use the --coreml flag to run inference with the custom v1.1.0 model featuring 39 merchandise-focused classes and 3.4x faster performance:

# On Apple Silicon (recommended)
artifactiq analyze --input photos/ --coreml

# Check if CoreML is available
artifactiq info

Performance Comparison

Backend	Model	Avg Inference	Classes
ONNX	v1.0.0	24ms	80 (COCO)
CoreML	v1.1.0	7ms	39 (Custom)

Custom Model Classes

Person, Man, Woman, Footwear, Clothing, Handbag, Hat, Sunglasses, Backpack, Jeans, Jacket, Coat, Dress, Camera, Mobile phone, Laptop, Watch, and more.

Note: The v1.1.0 custom model is currently CoreML-only (macOS/iOS). Cross-platform ONNX export is on the roadmap.

Install Options

# Install specific version
ARTIFACTIQ_VERSION=v1.0.0-alpha.13 curl -fsSL https://artifactiq.ai/install.sh | sh

# Install to custom directory
ARTIFACTIQ_INSTALL_DIR=/usr/local/bin curl -fsSL https://artifactiq.ai/install.sh | sh

Available Models

Use Axon to install detection models (requires Axon v3.3.0+ for ONNX conversion):

Model	Size	Speed	Accuracy	Install Command
yolov8n	6 MB	Fastest	Good	axon install hf/ultralytics/yolov8n --format onnx
yolov8s	22 MB	Fast	Better	axon install hf/ultralytics/yolov8s --format onnx
yolov8m	52 MB	Medium	Great	axon install hf/ultralytics/yolov8m --format onnx
yolov8l	87 MB	Slower	Excellent	axon install hf/ultralytics/yolov8l --format onnx
yolov8x	136 MB	Slowest	Best	axon install hf/ultralytics/yolov8x --format onnx

Usage Examples

# Basic detection
artifactiq analyze --input photo.jpg --model yolov8n

# JSON output for programmatic use
artifactiq analyze --input photo.jpg --model yolov8n --format json

# Process multiple images
artifactiq analyze --input ./images/ --model yolov8n

# With merchandise detection enabled
artifactiq analyze --input photo.jpg --model yolov8n --merchandise

# Set confidence threshold
artifactiq analyze --input photo.jpg --model yolov8n --confidence 0.5

CLI Reference

# Show help
artifactiq --help

# Show version and system info
artifactiq info

# List available/installed models
artifactiq models
artifactiq models --installed

# Download a model (via Axon)
artifactiq download --model yolov8n

# Analyze images
artifactiq analyze --input <path> --model <model>

Downloads

See Releases for all binaries.

Available Platforms

Platform	Architecture	File
Linux	x86_64	artifactiq-linux-amd64.tar.gz
macOS	Apple Silicon (arm64)	artifactiq-darwin-arm64.tar.gz

Manual Installation

# Download using gh CLI
gh release download --repo ARTIFACTIQ/releases --pattern "*arm64*"

# Or using curl (macOS Apple Silicon)
curl -LO https://github.com/ARTIFACTIQ/releases/releases/latest/download/artifactiq-darwin-arm64.tar.gz

# Or Linux x86_64
curl -LO https://github.com/ARTIFACTIQ/releases/releases/latest/download/artifactiq-linux-amd64.tar.gz

# Verify checksum
shasum -a 256 -c *.sha256

# Extract and install
tar xzf artifactiq-*.tar.gz
chmod +x artifactiq
sudo mv artifactiq /usr/local/bin/

Model Management with Axon

mlOS Axon handles model downloads, caching, and format conversions.

# Search for models
axon search yolo

# Install a model with ONNX conversion
axon install hf/ultralytics/yolov8n --format onnx

# List installed models
axon list

# Get model info
axon info yolov8n

# Update models
axon update

Verification

All releases include SHA256 checksums (.sha256 files) for verification:

shasum -a 256 -c artifactiq-*.sha256

Troubleshooting

"Model not found" error

Ensure Axon v3.3.0+ is installed and the model is downloaded:

# Check Axon installation (requires v3.3.0+ for YOLO ONNX)
axon --version

# Install the model with ONNX conversion
axon install hf/ultralytics/yolov8n --format onnx

# Verify model is installed
axon list

"Axon not configured" error

Install Axon first:

curl -sSL axon.mlosfoundation.org | sh

Then restart your terminal or run:

source ~/.bashrc  # or ~/.zshrc

Permission denied

Ensure the binary is executable:

chmod +x ~/.local/bin/artifactiq

Support

• Issues: GitHub Issues
• Contact: [email protected]
• Website: artifactiq.ai

Source Code

The source code is maintained in a private repository. For licensing inquiries, contact [email protected].