🦙
Llama
by Meta
Llama is Meta's family of open-source large language models designed for efficiency and performance. Developers deploy it on-premise or cloud for cost-effective AI inference.
Llama features and capabilities, including which plan they need and how they compare in the CanIUse matrix.
Used in combinations
Compare Llama with:
✨ Modalities
💬
Yes. Core featureText Input
🖼️
Yes. Llama 3.2Image Upload / Vision
🎨
NoImage Generation
🎙️
NoVoice Mode
🎥
NoVideo Understanding
🎬
No. Not a video generation product.Video Generation
📁
NoFile Upload (PDF, DOCX)
💻
NoCode Execution
🔊
NoAudio Input
🖌️
NoCanvas / Artifact Editor
🖥️
NoScreen Sharing / Computer Use
🧠 Memory & Context
🧠
NoPersistent Memory
📏
Yes. 128k tokensLarge Context Window
⚙️
YesCustom Instructions
📜
NoConversation History
🔌 Integrations
🌐
NoWeb Search / Browsing
🔌
NoPlugins / Extensions
🔧
Yes. Free via API providersAPI Access
⚡
NoZapier Integration
🤖
NoCustom GPTs / Bots
👥
NoReal-time Collaboration
🐦
Live Social Data Integration
📚
Live Citation Mapping
👨💻 Developer
🛠️
YesFunction Calling
⚡
YesStreaming Responses
🎛️
Yes. Open weightsFine-Tuning
📐
YesEmbeddings API
📋
YesStructured Output (JSON)
🧩
Partial. Llama 3.3 has some CoT capabilityReasoning / Thinking Mode
🏠
Yes. Weights are downloadable and run locally via llama.cpp, Ollama and similar runtimes.Local Execution
🔓
Open Weights
🤖
Multi-Agent Orchestration
📱 Platforms
🌐
NoWeb App
📱
NoiOS App
🤖
NoAndroid App
🖥️
NoDesktop App
🔵
NoChrome Extension
🔒 Privacy
🔒
YesGDPR Compliant
🚫
YesOpt Out of Data Training
🏢
Yes. Self-hostedEnterprise Privacy Mode
💼 Workspaces
📂
NoProjects / Folders