Meta · open weights · large
Meta: Llama 3.3 70B Instruct
Open weights, so you can run it locally. The published id includes 70B. That label is copied from the id, not measured here. Size class is this site's filter bucket, not a benchmark.
API price, in pounds
- Input, per million
- £0.529
- Output, per million
- £0.529
OpenRouter lists $0.7100 in and $0.7100 out per million tokens. Converted at 0.74501 on 2026-11-10. Fetched 2026-11-11. Context length listed as 131,072 tokens.
Sourced local speeds
No sourced tokens-per-second figure for this model. A speed is not invented. You can type one in the calculator if you have measured it.
Comparisons
- Meta: Llama 3.3 70B Instruct on GeForce RTX 5090
- Meta: Llama 3.3 70B Instruct on GeForce RTX 3090
- Meta: Llama 3.3 70B Instruct on GeForce RTX 4090
- Meta: Llama 3.3 70B Instruct on GeForce RTX 4070 Ti SUPER
- Meta: Llama 3.3 70B Instruct on GeForce RTX 4080 SUPER
- Meta: Llama 3.3 70B Instruct on GeForce RTX 5070 Ti