Cargo Registry Read the full architecture docs
Browse available cargos
A Cargo is a portable, signed unit of computation that runs on Islands across the Archipelag.io network.
Think of it as a shipping container for code — each one packages a model, runtime, and resource requirements into a single deployable artifact. Consumers submit jobs, the coordinator finds the best Island, and the Cargo executes in a secure sandbox.
Consumer
Cargo
Island
ONNXDockerWASMGGUF
Mistral 7B InstructOfficial
Mistral 7B Instruct v0.2 chat inference using llama.cpp
Qwen3.5-0.8BOfficial
Ultra-lightweight language model for edge and mobile inference
Qwen3.5-4BOfficial
Compact language model with strong reasoning capabilities
Qwen3.5-9BOfficial
Mid-size language model for consumer GPU inference
Qwen3.5-27BOfficial
High-quality language model for demanding reasoning tasks
Qwen3.5-35B-A3B (MoE)Official
Mixture-of-Experts LLM with 35B total / 3B active parameters
Qwen3-CoderOfficial
Code-specialized LLM for code generation, completion, and explanation
GLM-5Official
Massive MoE LLM for reasoning, coding, and agentic tasks
Llama 3.1 8BOfficial
Meta's general-purpose LLM with large fine-tune ecosystem
GPT-OSS 20BOfficial
OpenAI's first open-weight language model
MiniMax M2.5Official
Agentic LLM optimized for complex multi-step workflows
Nanbeige4.1-3BOfficial
Lightweight multilingual LLM with strong Chinese language support
Nemotron-3 120B-A12BOfficial
NVIDIA's MoE LLM with 120B total / 12B active parameters
Stable DiffusionOfficial
Text-to-image generation using Stable Diffusion 1.5 and SDXL
Music GenOfficial
Text-to-music generation using Meta MusicGen
FLUX.1-devOfficial
High-quality text-to-image generation with superior prompt adherence
FLUX.1-schnellOfficial
Ultra-fast text-to-image generation (4-step distilled)
Z-Image-TurboOfficial
Ultra-fast text-to-image generation from Alibaba
Stable Diffusion 3.5 LargeOfficial
Latest text-to-image from Stability AI with MMDiT architecture
Qwen-ImageOfficial
Multimodal image generation from Alibaba's Qwen family
Wan2.1-T2V-14BOfficial
High-quality text-to-video generation (14B parameters)
Wan2.2-T2V-A14BOfficial
Text-to-video with MoE architecture for efficient inference
HunyuanVideo-1.5Official
High-fidelity text-to-video generation from Tencent
Wan2.1-T2V-1.3BOfficial
Lightweight text-to-video for consumer GPUs
CogVideoX-2bOfficial
Efficient text-to-video generation from Z.AI
DistilBERT SentimentOfficial
Classifies text sentiment as positive or negative using DistilBERT
BART-CNN SummarizationOfficial
Summarizes long text using BART-large fine-tuned on CNN/DailyMail
OPUS-MT TranslationOfficial
Translates text between 100+ language pairs using MarianMT
BERT-NEROfficial
Identifies persons, organizations, and locations in text
DistilBERT QAOfficial
Extracts answers from context passages using DistilBERT
BERT Base Fill-MaskOfficial
Predicts masked tokens in text using BERT
BART-MNLI ClassificationOfficial
Classifies text into arbitrary categories using BART
T5 Grammar CorrectionOfficial
Corrects grammar in text using a fine-tuned T5 model
T5 ParaphrasingOfficial
Generates paraphrased versions of input text using T5
Toxic-BERTOfficial
Detects toxic content in text using a fine-tuned BERT classifier
KeyBERT ExtractionOfficial
Extracts keywords and keyphrases from text using KeyBERT
MiniLM-L6 EmbeddingsOfficial
Generates text embeddings using sentence-transformers
CodeGen-350MOfficial
Generates code completions using CodeGen for Python
YOLOv8 DetectionOfficial
Detects objects in images with bounding boxes using YOLOv8
YOLOv8 SegmentationOfficial
Performs instance segmentation on images using YOLOv8-seg
BLIP CaptioningOfficial
Generates natural language descriptions of images using BLIP
DPT-Hybrid DepthOfficial
Estimates depth maps from single images using DPT-Hybrid
MTCNN-FaceNetOfficial
Detects faces and generates facial embeddings using MTCNN + FaceNet
U2-Net Background RemovalOfficial
Removes background from images, returns transparent PNG
Style TransferOfficial
Applies artistic style transfer to images
Real-ESRGAN UpscalingOfficial
Upscales images 2x or 4x using Real-ESRGAN
GLM-OCROfficial
Advanced document understanding and OCR using GLM vision model
NuMarkdownOfficial
Converts document images to structured Markdown with reasoning
FireRed-OCROfficial
High-accuracy OCR for documents, handwriting, and scene text
Whisper BaseOfficial
Transcribes and translates audio with word-level timestamps
Tacotron2 TTSOfficial
Converts text to speech audio using Coqui TTS
Whisper Large-v3-TurboOfficial
Fast speech-to-text (2x faster than large-v3, near-identical quality)
Qwen3-ASROfficial
Multilingual speech recognition from Qwen family
Voxtral-ASROfficial
Realtime streaming speech recognition from Mistral AI
VibeVoice-ASROfficial
High-accuracy speech recognition from Microsoft
Parakeet-ASROfficial
Fast accurate speech recognition from NVIDIA (0.6B parameters)
Kokoro TTSOfficial
Ultra-lightweight high-quality text-to-speech (82M parameters)
Qwen3-TTSOfficial
Voice cloning and custom voice text-to-speech from Qwen family
Chatterbox TTSOfficial
High-quality voice cloning text-to-speech
VibeVoice TTSOfficial
Multi-speaker instruction-following text-to-speech from Microsoft
CSM TTSOfficial
Conversational speech model optimized for natural dialogue
IndexTTS-2Official
High-quality multilingual text-to-speech
VibeVoice RealtimeOfficial
Realtime streaming text-to-speech optimized for low latency
Fish Audio S2Official
Multi-speaker text-to-speech with instruction-following control
Qwen3-Embedding-0.6BOfficial
Lightweight text embeddings from Qwen family
Qwen3-Embedding-8BOfficial
High-quality text embeddings for RAG from Qwen family
E5-Large MultilingualOfficial
Battle-tested multilingual text embeddings
BGE-SmallOfficial
Fast lightweight English text embeddings
Jina Embeddings v5Official
Latest-generation text embeddings from Jina AI
Qwen3-VL-EmbeddingOfficial
Vision+text multimodal embeddings from Qwen family
Video TranscodeOfficial
Video transcoding and format conversion
Video CompressOfficial
Video compression and transcoding
Video ClipOfficial
Extract a time range from a video
Video ThumbnailOfficial
Extract thumbnails from video
Video SubtitleOfficial
Video subtitling using Whisper
GIF MakerOfficial
GIF creation from images or video
ScreenshotOfficial
Capture screenshots from URLs using headless Chrome
Audio ConvertOfficial
Audio format conversion using ffmpeg
Audio MergeOfficial
Merge and concatenate audio files
Audio SplitOfficial
Split audio into segments at silence boundaries
Noise RemoveOfficial
Audio noise removal using noisereduce
Normalize AudioOfficial
Normalize audio volume levels
OCROfficial
Optical character recognition for image-to-text
PDF ExtractOfficial
PDF text and content extraction
PDF MergeOfficial
PDF merging and splitting
PDF SignOfficial
Add text watermark or signature to PDF pages
HTML to PDFOfficial
Convert HTML to PDF using WeasyPrint
Markdown to PDFOfficial
Convert Markdown to styled PDF
DOCX ExtractOfficial
Extract text and images from Word documents
Excel ParseOfficial
Parse Excel spreadsheets to JSON
Batch TemplateOfficial
Generic data processing template for batch jobs
ResizeOfficial
Image resizing and thumbnail generation
Compress ImageOfficial
Image compression with quality control
Convert ImageOfficial
Image format conversion
Smart CropOfficial
Smart crop using edge-based saliency detection
Blur FaceOfficial
Blur faces in images for privacy
WatermarkOfficial
Add text or image watermarks
ColorizeOfficial
Colorize grayscale images
Image DiffOfficial
Compare two images and highlight differences
Image MetadataOfficial
Extract and manipulate image EXIF metadata
SVG OptimizeOfficial
SVG optimization and minification
SVG to PNGOfficial
Convert SVG to PNG using CairoSVG
BarcodeOfficial
Generate and read barcodes
Format CodeOfficial
Code formatting for multiple languages
JSON to TypesOfficial
Generate TypeScript or Python types from JSON
Language DetectOfficial
Language detection using lingua
LLM Chat (Mock)Official
Mock LLM chat for testing
Stable Diffusion (Mock)Official
Mock image generation for testing
Qwen3.5-0.8B (CoreML)Official
Ultra-lightweight on-device LLM for iPhone Neural Engine
WhisperKit (CoreML)Official
On-device speech recognition using pre-compiled CoreML Whisper
Kokoro TTS (CoreML)Official
Ultra-lightweight on-device text-to-speech for iOS
HashOfficial
SHA-256, MD5, and Blake3 hash computation
JSONOfficial
JSON processing with jq-like operations
CSVOfficial
CSV processing with JSON conversion and filtering
Base64Official
Base64 encode and decode operations
CompressOfficial
Gzip compression and decompression
RegexOfficial
Regex testing and extraction
MarkdownOfficial
Markdown to HTML conversion (GFM)
Markdown TOCOfficial
Generate table of contents from Markdown
UUIDOfficial
UUID generation (v4, v7)
QR CodeOfficial
QR code generation from text or URLs
JWTOfficial
JWT decode and inspect
SanitizeOfficial
HTML sanitization for XSS protection
YAMLOfficial
YAML to JSON conversion
TOMLOfficial
TOML to JSON conversion
URLOfficial
URL parsing and manipulation
SlugOfficial
URL slug generation
SemverOfficial
Semantic version parsing and comparison
DiffOfficial
Text diff generation
HighlightOfficial
Syntax highlighting with HTML output
MinifyOfficial
HTML, CSS, and JS minification
CronOfficial
Cron expression parsing and next-run calculation
EchoOfficial
Test cargo that echoes JSON input
Submit your own cargo
Package your model, container, or WASM module as a cargo and publish it to the Archipelag.io network. Community cargos start unverified — pass our security review to earn the Certified badge and unlock higher trust levels.
CommunitySubmit a cargo — runs sandboxed, no network access
CertifiedPass security review — higher resource limits, trusted by consumers
OfficialMaintained by Archipelag.io — full network and GPU access