Back to Models
Llama 3.1 405B
Metatext9.2 / 10 Overall Rating
The largest open-weight model — competitive with proprietary leaders.
Context Window
128K
Knowledge Cutoff
2025
Max Output
16K
License
Open Weights
Where it Excels (Pros)
- Fully open-source
- Competitive benchmarks
- Self-hostable
- No usage fees
Limitations (Cons)
- Requires significant compute to run
- Slightly behind GPT-4o/Claude on hard tasks
- Less polished instruction following
Benchmark Breakdown
Reasoning (MMLU)88%
Coding (HumanEval)86%
Mathematics (MATH)84%
Long-Context Retrieval82%
Pricing Matrix
Input cost / 1M tokens$0.80
Output cost / 1M tokens$0.80
Compare Llama 3.1 405B with other frontier models: