// Open-model VRAM table · family
Llama: local deployment & VRAM
16 open-weight models from Meta (34 official repos incl. quantised builds). At 4-bit they range from 1.6 GB (Llama-3.2-1B-Instruct) to 275 GB (Llama-3.1-405B).
official orgs: meta-llamalatest 2025-04-28memory = estimate
01Text LLM (6)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| Llama-3.3-70B-Instruct | 70.6B | 128K | 158 GB | 50.2 GB | 2 × RTX 5090 (64GB) | Mac 128GB | Commercial with conditions | 2024-11-26 |
| Llama-3.2-1B-Instruct | 1.24B | 128K | 3.5 GB | 1.6 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2024-09-18 |
| Llama-3.2-3B-Instruct | 3.21B | 128K | 8.5 GB | 3.6 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2024-09-18 |
| Llama-3.1-8B-Instruct | 8.03B | 128K | 19.2 GB | 6.9 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2024-07-18 |
| Llama-3.1-70B-Instruct | 70.6B | 128K | 158 GB | 50.2 GB | 2 × RTX 5090 (64GB) | Mac 128GB | Commercial with conditions | 2024-07-16 |
| Llama-3.1-405B | 405.9B | 128K | 898 GB | 275 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Commercial with conditions | 2024-07-16 |
02Vision-language (3)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| Llama-4-Scout-17B-16E-Instruct | 108.6B | — | 240 GB | 73.0 GB | RTX PRO 6000 96GB | Mac 128GB | Commercial with conditions | 2025-04-02 |
| Llama-4-Maverick-17B-128E-Instruct | 401.6B | — | 884 GB | 268 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB | Commercial with conditions | 2025-04-01 |
| Llama-3.2-11B-Vision-Instruct | 10.7B | — | 24.0 GB | 7.6 GB | 12GB GPU | Mac 16GB | Commercial with conditions | 2024-09-18 |
03Safety / guard (7)
| Model | Params | Context | BF16 | 4-bit | Smallest GPU (4-bit) | Mac (4-bit) | Licence | Released |
|---|---|---|---|---|---|---|---|---|
| Llama-Prompt-Guard-2-22M | 71M | — | 0.7 GB | 0.5 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2025-04-28 |
| Llama-Prompt-Guard-2-86M | 279M | — | 1.1 GB | 0.7 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2025-04-28 |
| Llama-Guard-4-12B | 12B | — | 26.9 GB | 8.5 GB | 12GB GPU | Mac 16GB | Commercial with conditions | 2025-04-23 |
| Llama-Guard-3-1B | 1.5B | — | 3.8 GB | 1.5 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2024-09-20 |
| Llama-Guard-3-11B-Vision ↗ | 10.7B | — | 24.0 GB | 7.6 GB | 12GB GPU | Mac 16GB | Commercial with conditions | 2024-09-20 |
| Llama-Guard-3-8B | 8.03B | — | 18.2 GB | 5.9 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2024-07-22 |
| Prompt-Guard-86M | 279M | — | 1.1 GB | 0.7 GB | 8GB GPU | Mac 16GB | Commercial with conditions | 2024-07-21 |
04Other families
Qwen · IBM Granite · NVIDIA Nemotron · Gemma · Mistral / Magistral / Devstral · GLM / CogVideoX (Z.ai) · DeepSeek · Ai2 OLMo / Molmo · Ant Ling / Ring / Ming · Liquid LFM · MiniCPM / VoxCPM · Tencent Hunyuan / Hy · Phi · Meituan LongCat · Jina embeddings & rerankers · Wan · Kimi · Cohere Command / Aya · Xiaomi MiMo · FLUX · Microsoft VibeVoice / Fara / Harrier · InternLM / Intern-S · StepFun Step · MiniMax · Baidu ERNIE · NVIDIA Parakeet / Canary speech · Stable Diffusion / Stable Audio · Yi (01.AI) · ByteDance Seed · SmolLM / SmolVLM · BGE embeddings & rerankers · LTX Video · FunAudioLLM speech · gpt-oss · poolside Laguna · Whisper · Baichuan · Thinking Machines Inkling · xAI Grok · rednote dots
Totals include 10% + 0.5GB overhead and one 8K-token KV cache for text models. Estimates only. Method →