Llama
Featured
Platforms & Models
Meta
FreeOpen Source
About
Meta's open-source LLM series, deployable locally with commercial use allowed
Key Features
Llama 4 natively multimodal models
Scout, Maverick and Behemoth variants
Up to 10M token context (Scout)
MoE architecture, 200-language pretraining
Open weights, commercial use
Local and cloud deployment
Pricing
free
Open weights, free to use
cloud
Hosted API via providers (pay-as-you-go)
hardware
Local deployment needs GPU
Use Cases
Building custom AI applications
Local and private LLM deployment
Fine-tuning for domains
Multilingual assistants
Research and experimentation
Cost-efficient inference
Pros
Open weights, commercial use
Strong multilingual performance
Very long context (Scout)
Large ecosystem support
Cons
Large models need heavy hardware
License has some usage limits
Setup requires expertise
Latest Update
July 2026: Meta's Llama 4 herd (Scout, Maverick, Behemoth) brought native multimodality and up to 10M token context