Llama 3.1-405B
Llama 3.1-405B is an open-weights language model from Meta, released July 23, 2024. It has 406B parameters (dense). On the Epoch Capabilities Index it scores 127.8, rank 162 of 256 rated models.
Check it against your own PC →
Can you run Llama 3.1-405B locally?
Even with 32 GB of VRAM and 32 GB of RAM it is too large to run at a usable speed on a single consumer GPU. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Too large | – | – | – | – |
| 8 GB VRAM | Too large | – | – | – | – |
| 12 GB VRAM | Too large | – | – | – | – |
| 16 GB VRAM | Too large | – | – | – | – |
| 24 GB VRAM | Too large | – | – | – | – |
| 32 GB VRAM | Too large | – | – | – | – |
Ready-made quantizations on Hugging Face: ThomasBaruzier/Meta-Llama-3.1-405B-Instruct-GGUF · clowman/Llama-3.1-405B-Instruct-AWQ-Int4 · NousResearch/Hermes-4-405B-FP8 · genevaquant/Hermes-4-405B-FP8 · rax4901/Hermes-4-405B-FP8 · RedHatAI/Meta-Llama-3.1-405B-FP8
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Math | 14.0 | MATH level 5, OTIS Mock AIME 2024-2025 |
| Knowledge | 35.6 | GPQA diamond, MMLU |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| MMLU | 79.3% |
| MATH level 5 | 49.8% |
| GPQA Diamond | 36.0% |
| OTIS Mock AIME | 9.6% |
| SimpleBench | 7.6% |
| The Agent Company | 7.4% |
Links
Hugging Face: meta-llama/Llama-3.1-405B · Epoch AI benchmarks
Models with a similar score
DeepSeek-V3 · Qwen2.5-Max · Gemini 1.5 Pro (Sept 2024) · Claude 3.5 Sonnet · Grok-2 (Dec 2024) · Phi-4 · GPT-4o (Nov 2024) · GPT-4o (May 2024)
More from Meta
Muse Ember · Muse Spark Open · Muse Wonder · Muse Wonder Full · Muse Spark 1.3 · Muse Spark 1.3 Contributor · Muse Glimmer · Muse Glimmer 30B assistant