A family of open models for everyday conversations, reasoning, coding, and multimodal tasks.

Google’s compact open models for local conversations, writing, and multimodal tasks.

Meta’s lightweight open models for conversations, rewriting, summarization, and everyday assistance.

Compact Mistral models for local conversations, instruction following, and image understanding.

IBM’s open models for structured writing, document work, coding assistance, and professional tasks.

Microsoft’s compact models for writing, coding help, summarization, and structured tasks.

Liquid AI’s efficient models for responsive on-device chat, reasoning, and lightweight image understanding.

Oscilla’s lightweight local model for fast, private conversations and everyday assistance on Apple devices.

Hugging Face’s compact multilingual model for efficient conversations, writing, and local reasoning.

OpenBMB's compact English and Chinese text models for local chat, coding, and problem-solving.

A specialized Vision family for screenshots, images, scanned documents, charts, and visual question answering.

NVIDIA's compact instruction family for detailed answers, technical questions, and structured problem-solving.

Prism ML's efficiency-focused family, ranging from extremely lightweight chat to higher-quality local reasoning.