— Concepts
Open-Weights Model
An LLM whose trained parameters are publicly downloadable — Llama, Mistral, DeepSeek, Sarvam.
Also known as: Open weights · Open-source LLM
What is Open-Weights Model?
An open-weights model is one whose trained parameter weights are publicly downloadable — letting you run it on your own hardware, fine-tune it, and deploy it without per-token vendor fees. Distinct from 'open-source' (which would also include training data and code, rare in practice). Leading 2026 open-weights families: Llama (Meta), Mistral, DeepSeek, Qwen, Sarvam.
— Related
Terms connected to Open-Weights Model
Models
LLM (Large Language Model)
An AI model trained on huge amounts of text that can understand and generate human language.
Open →Techniques
Fine-Tuning
Adjusting a pre-trained AI model on your specific data to change its behaviour.
Open →Models
DeepSeek V4
Chinese open-weight LLM — very strong reasoning at a fraction of Opus pricing.
Open →Models
Mistral Small 4
Compact European open-weight model — strong on European languages, runs on a single GPU.
Open →India
Sarvam AI
Bengaluru-based Indic LLM lab — open-source 105B model, multilingual TTS/STT, India-first.
Open →From definitions to deployed projects.
Knowing what a term means is step one. ONROL's AI Generalist track gets you shipping projects that use it.