Llama
Meta's open-weight family spanning Llama 3.1 (8B–405B) and Llama 3.2 (1B–90B vision)
Free and open-weight, API via Together AI, Fireworks, Groq
Overview
Llama is Meta's family of open-weight large language models, released for research and commercial use. Llama 3 models are among the most capable open-source LLMs available, and are the foundation for thousands of fine-tuned models and local AI applications worldwide.
Key Features
- Fully open weights — run locally or self-host at scale
- Models ranging from 1B (mobile/edge) to 405B (frontier performance)
- Llama 3.1 and 3.2 offer strong coding, reasoning, and instruction following
- Multilingual support for 8 languages in Llama 3.1
- Vision models for image understanding (Llama 3.2)
- Commercial license permitting use in products with up to 700M monthly users
Pricing: Free and open-source; infrastructure costs vary by deployment method.
Pros
- Fully open-weight with permissive commercial licensing
- Sizes from 1B on-device to 405B frontier-class
- Massive ecosystem across all major cloud and inference providers
- Self-hosting eliminates per-token costs at scale
Cons
- Large models require significant GPU infrastructure
- License restricts use above 700M monthly active users
- Trails top proprietary models on complex reasoning tasks
Tags
Product Updates
Similar Tools
Qwen
Alibaba's comprehensive open-weight family spanning text, code, vision, audio, and speech
DeepSeek
Chinese AI lab's open-weight models with frontier reasoning at ultra-low cost
GLM (Zhipu AI)
Chinese AI lab with open-weight models spanning text, vision, voice, and video generation
Kimi
Moonshot AI's frontier open-source model family, led by Kimi k2.5 with exceptional coding and long-context reasoning
