by meta-llama
Llama 3.3 70B Instruct
meta-llama/Llama-3.3-70B-Instruct
All figures are estimates. Real memory use and speed depend on your drivers, settings and other programs running.
- Parameters
- 71B
- Licence
- llama3.3
- Published
- Nov 26, 2024
The licence is the model maker’s. Check its terms before you use the model.
Hugging Face asks you to accept the maker’s terms before you download this model. No kit is sold for it.
Original on Hugging Face ↗Memory needed by quantization
At 8,192 tokens of context, including 0.8 GB for the runtime.
| Quantization | Model file | Total |
|---|---|---|
| Q4_K_M | 42.8 GB | 47.9 GB(rough) |
| Q5_K_M | 50.2 GB | 56.0 GB(rough) |
| Q6_K | 57.9 GB | 64.4 GB(rough) |
| Q8_0 | 75.0 GB | 83.3 GB(rough) |
| F16 | 141.1 GB | 156.0 GB(rough) |
Hardware that runs it on the graphics card
At Q4_K_M and 8,192 tokens of context.
None of the hardware we list runs it fully on the graphics card at this setting.
Check it on your computerThe same model through an API
Cheapest seller: NanoGPT, $0.99 a month at the default use (200 requests a day).
See the prices on the Model Price MapModel details come from Hugging Face’s public listings (names, sizes, licences). Every model page links to the original.