Llama 3.1 8B Instruct vs Llama 3.2 1B Instruct
Llama 3.2 1B Instruct is the cheaper option on input by 46%; Llama 3.1 8B Instruct is cheaper on output by 60%. Figures below are per million tokens, from vendor pricing cited in our reporting.
| Llama 3.1 8B Instruct | Llama 3.2 1B Instruct | |
|---|---|---|
| Input per 1M tokens | $0.05 | $0.03 |
| Output per 1M tokens | $0.08 | $0.20 |
| Cached input per 1M | $0.03 | — |
| Context window | 131K tokens | 60K tokens |
| Parameters | — | — |
| Licence | open weights | open weights |
| Benchmark | — | — |
| Vendor | Meta | Meta |
| Released | — | — |
What the numbers say
- Input price gap: 46% ($0.05 vs $0.03).
- Blended three-to-one input/output rate: $0.23 for Llama 3.1 8B Instruct, $0.28 for Llama 3.2 1B Instruct.
- Context: 131K tokens against 60K tokens.
- Licence: Llama 3.1 8B Instruct is open weights; Llama 3.2 1B Instruct is open weights.
Sources
Llama 3.1 8B Instruct specs · Llama 3.2 1B Instruct specs · Full price index