Small AI Models (Under 3B)

Tiny embedding model. Only 23MB. Perfect for on-device search.

Compact English embedding model. Good for basic semantic search.

High quality text embedding model. 137M params. Good for RAG and search.

Compact 1.5B model with strong multilingual and coding abilities.

High quality English embedding model. Best accuracy for English search.

Optimized large Whisper model. Near-best accuracy with faster inference.

Multilingual reranker. 100+ languages. 1.1GB.

Ultra-small 0.5B model from Alibaba. Minimal resource requirements.

Largest Whisper model. Best accuracy across all languages and accents.

Ultra-compact 1B model. Runs on virtually any device including smartphones.

Lightweight 1.1B chat model based on Llama architecture. Great for phones.

Ultra-compact vision model. Only 1GB. Answers questions about images.

Compact vision-language model. Default multimodal model. Can understand images and answer questions about them.

Compact Whisper model. Good accuracy for everyday transcription tasks.

Classic image generation model. Pre-converted to CoreML for iOS/Mac. Downloads as zip, auto-extracts.

Base whisper model. Good balance of speed and accuracy. 142MB.

Distilled Whisper. 6x faster than large-v3 with 1% accuracy loss.

Tiny 135M model. Default LLM - guaranteed to run on any iPhone. Only 145MB download. Perfect for quick experiments.

Google's latest tiny 1B model. Excellent quality for its size.

Tiny multilingual speech recognition. Only 75MB. Supports 99 languages. Runs on any device.

Compact reasoning model distilled from DeepSeek R1. Strong chain-of-thought in a tiny package.

Compact code model with solid code generation and understanding abilities.

Mid-size Whisper model. Strong multilingual speech recognition.

Compact 360M model. Good for basic tasks on very constrained devices.

Smallest code model. Default code assistant - runs on any iPhone. Great for code completion and simple programming tasks.

Google's compact 2.6B model. Efficient and capable for mobile use.

Efficient multimodal model with strong image understanding. Optimized for edge devices.

Capable 1.7B model from HuggingFace. Good balance for mobile devices.

Music generation from text prompts. Requires multiple ONNX files (~435MB total). Experimental iOS support.

Smallest possible speech recognition model. Only 32MB. English only. Default speech model - guaranteed to run on any iPhone.

High quality 82M parameter TTS model. Excellent speech synthesis with multiple voice options. 86MB download.

English-only base model. Faster and more accurate for English.

Compact code model with strong coding capabilities. Great for mobile coding assistants.

Compact model from LG. Optimized for Korean and English.

Compact embedding model from Snowflake. Good multilingual support.

Ultra-tiny 500M model. Even smaller than SmolLM. Runs anywhere.

IBM's compact 2B model. Good at following instructions.

Lightweight code completion model from Google. Fast on-device code suggestions.

Ultra-compact 1B model from Technology Innovation Institute.

SD 3 with MMDiT architecture. Superior text rendering.

Tiny English reranker. Only 67MB. Use with embedding models for better search.

SD 1.5 in single-file GGUF format. Alternative to CoreML. Uses stable-diffusion.cpp with Metal acceleration.

Tiny code model. Great for phones. Fast completions.

Smallest CoreML image generation model. Palettized for minimal size (1.14GB). Runs on any iPhone with 6GB RAM. Default image generation model.

Lightweight TTS voice. High quality English speech synthesis. Default TTS model - runs on any iPhone. Only 63MB.

High quality English male voice. 63MB download. Runs on any device.

Medium quality English voice with natural prosody. 63MB download.

SD 2.1 in GGUF format. Better quality than 1.5.

Spanish female voice. Natural prosody.

French female voice.

German male voice.

Chinese Mandarin voice.

Japanese voice.

Korean voice.

Russian female voice.

Portuguese voice.

Italian male voice.

Arabic voice.