← Back to directory
Meta · Released 2025-09
Llama 4
Meta's flagship dense LLM. Comes in 8B, 70B, and 405B; 8B fits on a single 16GB GPU and is the most-downloaded local model on Hugging Face.
LlamaCommercial with caveats👁 Visiongeneralvision
Params (max)
405B
Variants
405B / 70B / 8B
Context window
128K tokens
MMLU
87.1
HumanEval
84.5
GSM8K
93
Min VRAM (fp16, smallest variant)
16GB
Smallest Q4 GGUF
~5GB
Languages supported
12
Pros
- ✓Massive ecosystem
- ✓Strong instruction following
- ✓Fine-tunes everywhere
Cons
- ×License restricts >700M MAU products
- ×405B variant heavy for self-hosting
Highlights
- ●8B variant runs on a laptop
- ●Vision input on 90B+
- ●Most ecosystem support of any open LLM
Where to download
Hugging Face: meta-llama/Meta-Llama-4-405B
Or via Ollama (
ollama pull llama-4) or LM Studio's in-app browser.Homepage: https://llama.meta.com
Related reading
Best Open Source LLMs 2026: Honest Picks by Use Case
Which open-source LLM should you actually run in 2026? Honest picks by use case — frontier reasoning, coding, RAG, edge devices, multilingual.
Open Source LLM Licenses Explained: Llama vs Apache vs Gemma vs MIT
Can you use Llama in a commercial product? What does the Gemma license actually restrict? A plain-English breakdown of every major open LLM license.
DeepSeek V4 vs Llama 4: Which Open Frontier Model Should You Run?
DeepSeek V4 just topped the open leaderboard. Should you switch from Llama 4 405B? Side-by-side on benchmarks, license, hardware, and ecosystem.
Open Source LLMs vs Claude / GPT in 2026: When Does Open Win?
Open-source LLMs caught up to GPT-4 in 2024 and Claude Opus in 2026 — but should you actually switch? Cost, quality, latency, privacy compared.