Llama 3.1 405B
Open Weights
Meta
Meta's largest and most capable open-weights model, rivaling top proprietary models. First openly available model competitive with GPT-4, Claude 3 Opus, and Gemini 1.5 Pro. Trained on over 16 trillion tokens using 16,000+ H100 GPUs. Features 128K context window (16x larger than Llama 3) and supports 8 languages. 87.3% on general knowledge benchmark exceeds GPT-4 Turbo (86.5%) and Claude 3 Opus (86.8%). Enables commercial use with permissive license, allowing self-hosting and fine-tuning.
Strengths
- First open-weights model rivaling GPT-4, Claude 3 Opus, and Gemini 1.5 Pro
- 87.3% general knowledge score exceeds GPT-4 Turbo (86.5%) and Claude 3 Opus (86.8%)
- Permissive license allows commercial use, self-hosting, and fine-tuning
- 128K context window (16x larger than Llama 3) handles long documents
- Supports 8 languages: English, German, French, Italian, Portuguese, Hindi, Spanish, Thai
- Trained on 16T+ tokens using 16,000 H100 GPUs (unprecedented scale)
- No API costs if self-hosted - only infrastructure expenses
Caveats
- 405B parameters require significant GPU memory (8x A100 80GB minimum recommended)
- Self-hosting costs can be high for production-scale deployments
- No vision or audio capabilities (text-only)
- API provider pricing varies widely ($0.50-$5.00 per 1M tokens)
- Quantization required for consumer hardware reduces quality
- 4K max output tokens lower than some proprietary models
- Vendor
- Meta
- Context Window
- 128,000 tokens
- Max Output
- 4,096 tokens
- License
- Llama 3.1 Community License
- Release Date
- 2024-07-23
- Modalities
- text
| Benchmark | Score |
|---|---|
| General Knowledge | 87.3 |
Capabilities
Vision
Audio
Video
Tool Use
Pricing
Open weights - free to download and self-host. API providers vary ($0.50-$5.00 per 1M tokens typical)