Expired50% confidenceBenchmarkExact time
Based on the MLX framework, top-configured Mac models can achieve a generation speed of approximately 20 tokens/second when running the Llama-3-70B quantized model.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/18/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/16/2026(expired)
Sources
Related Claims
UnverifiedIn a controlled test, the 35B model ran at approximately 100 tokens/second (Mac Studio), roughly five times the 9B model's 16 tokens/second81% similarUnverifiedUnder a 4×RTX 3090 SXM4 configuration, the quantized version of GLM-5.2 achieves a token generation speed of approximately 6.3–6.8 tokens/second and a prompt evaluation speed of approximately 10.6 tokens/second.74% similarUnverifiedThe 4-bit quantized GGUF format version of llama.cpp enables the Llama 3.1 70B model to run on an Apple M2 Ultra with 64GB unified memory at an inference speed of approximately 20–30 tokens/second.74% similarUnverifiedApple M2/M3 chips adopt a Unified Memory Architecture (UMA), and llama.cpp can achieve approximately 30–50 tokens/second inference speed for 7B models via the Metal API.72% similarUnverifiedThrough pure CPU inference and the 4-bit GGUF quantization format, llama.cpp enables a 7B-parameter model to achieve a generation speed of 20–40 tokens per second on an ordinary server without a GPU.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/554308API
curl https://kongchang.com/api/v1/knowledge/claims/554308MCP
get_claim(id=554308)