readygpu

Llama 3.3 70B

70.6B parameters · 128k context · llama3.3 · released 2024-11-26

chatcoding

What to run it on in the UK

Fully in memory at 64k context, priced at retailers today.

Something we price holds it fully in memory, but none reaches 15 tok/s, our bar for recommending it. The table below has the speeds.

All hardware that runs it

Everything we price, best value first. Two cards split the model by layers.

HardwareQuantFitsDecodePricetok/s per 1,000
Mac Studio M4 Max 128GBQ8_0fits4 tok/s––
Framework Desktop Ryzen AI Max+ 395 128GBQ8_0fits2 tok/s––
Mac Studio M3 Ultra 512GBQ8_0fits6 tok/s––
NVIDIA DGX Spark 128GBQ8_0fits2 tok/s––
GMKtec EVO-X2 Ryzen AI Max+ 395 128GBQ8_0fits2 tok/s––
Mac Studio M3 Ultra 256GBQ8_0fits6 tok/s––
Mac Studio M3 Ultra 96GBQ5_K_Mfits8 tok/s––
NVIDIA RTX PRO 6000 BlackwellQ8_0fits12 tok/s––
ASUS Ascent GX10 128GBQ8_0fits2 tok/s––

31 other cards and machines we price cannot hold it at this context.

Memory needed

Weights, KV cache and runtime buffers at each context length (KV cache at fp16).

QuantFile@8k@32k@64k@128k
Q4_K_Mgguf39.6 GB42.8 GB50.5 GB60.7 GB81.1 GB
mlx-4bitmlx37.0 GB40.2 GB47.9 GB58.0 GB78.4 GB
Q5_K_Mgguf46.5 GB49.8 GB57.4 GB67.6 GB88.0 GB
mlx-8bitmlx69.8 GB73.1 GB80.7 GB90.9 GB111.3 GB
Q8_0gguf69.8 GB73.1 GB80.7 GB90.9 GB111.3 GB

Run it

Copy a command for the runtime you use.

  • Q4_K_M
    • Ollamaollama run hf.co/unsloth/Llama-3.3-70B-Instruct-GGUF:Q4_K_M
    • LM Studiolms get unsloth/Llama-3.3-70B-Instruct-GGUF@q4_k_m
    • llama.cppllama-server -hf unsloth/Llama-3.3-70B-Instruct-GGUF:Q4_K_M
  • mlx-4bit
    • LM Studiolms get mlx-community/Llama-3.3-70B-Instruct-4bit
    • MLXmlx_lm.server --model mlx-community/Llama-3.3-70B-Instruct-4bit
  • Q5_K_M
    • Ollamaollama run hf.co/unsloth/Llama-3.3-70B-Instruct-GGUF:Q5_K_M
    • LM Studiolms get unsloth/Llama-3.3-70B-Instruct-GGUF@q5_k_m
    • llama.cppllama-server -hf unsloth/Llama-3.3-70B-Instruct-GGUF:Q5_K_M
  • mlx-8bit
    • LM Studiolms get mlx-community/Llama-3.3-70B-Instruct-8bit
    • MLXmlx_lm.server --model mlx-community/Llama-3.3-70B-Instruct-8bit
  • Q8_0
    • Ollamaollama run hf.co/unsloth/Llama-3.3-70B-Instruct-GGUF:Q8_0
    • LM Studiolms get unsloth/Llama-3.3-70B-Instruct-GGUF@q8_0
    • llama.cppllama-server -hf unsloth/Llama-3.3-70B-Instruct-GGUF:Q8_0