Apple M1

Other

39

Run records

14

Models covered

4

Frameworks covered

L2 Cross-framework verified × 39

Evidence mix

Typical measured performance

Published benchmark averages per model × framework × quantization on this hardware.

  1. llama.cpp · IQ2_M Decode 88.05 · Prefill 311.52 ·
  2. llama.cpp · Q4_K_M Decode 55.57 · Prefill 194.52 ·
  3. Strata · Q4_K_S Decode 87.2 · Prefill 469.35 ·
  4. llama.cpp · Q3_K_M Decode 51.89 · Prefill 170.87 ·
  5. Strata · IQ1_S Decode 96.89 · Prefill 527.4 ·
  6. vLLM · Q4_K_S Decode 132.87 · Prefill 708.61 ·
  7. Strata · Q2_K Decode 231.36 · Prefill 1029.89 ·
  8. Ollama · IQ3_XS Decode 50.37 · Prefill 163.76 ·
  9. Ollama · Q4_K_M Decode 39.07 · Prefill 131.6 ·
  10. Ollama · IQ2_XS Decode 64.59 · Prefill 227.22 ·
  11. llama.cpp · Q3_K_L Decode 65.56 · Prefill 344.27 ·
  12. llama.cpp · Q8_0 Decode 36.81 · Prefill 167.78 ·
  13. Ollama · Q2_K Decode 72.5 · Prefill 393.97 ·
  14. llama.cpp · IQ3_S Decode 97.72 · Prefill 393.93 ·
  15. llama.cpp · Q5_K_M Decode 75.84 · Prefill 355.79 ·
  16. vLLM · Q4_K_M Decode 84.3 · Prefill 383.2 ·
  17. vLLM · Q6_K Decode 72.53 · Prefill 321.64 ·
  18. llama.cpp · IQ2_XS Decode 75.23 · Prefill 381.59 ·
  19. llama.cpp · Q4_K_M Decode 59.36 · Prefill 212.58 ·
  20. vLLM · FP16 Decode 30.09 · Prefill 100.41 ·
  21. vLLM · IQ2_XS Decode 79.87 · Prefill 436.11 ·
  22. vLLM · IQ4_XS Decode 47.87 · Prefill 181.96 ·
  23. Ollama · FP16 Decode 21.35 · Prefill 72.87 ·
  24. Ollama · Q3_K_L Decode 41.49 · Prefill 134.08 ·
  25. llama.cpp · IQ2_M Decode 189.83 · Prefill 1001.05 ·
  26. llama.cpp · Q2_K Decode 189.43 · Prefill 860.24 ·
  27. vLLM · Q3_K_S Decode 170.33 · Prefill 911.66 ·
  28. vLLM · Q4_K_M Decode 143.97 · Prefill 610.72 ·
  29. Ollama · IQ3_XS Decode 74.16 · Prefill 283.84 ·
  30. Ollama · Q6_K Decode 48.37 · Prefill 198.04 ·
  31. Strata · Q3_K_M Decode 114.61 · Prefill 400.65 ·
  32. llama.cpp · IQ4_XS Decode 148.84 · Prefill 610.28 ·
  33. Ollama · Q2_K Decode 166.63 · Prefill 805.73 ·
  34. llama.cpp · IQ2_XS Decode 125.44 · Prefill 495.34 ·
  35. llama.cpp · Q6_K Decode 62.45 · Prefill 267.16 ·
  36. Ollama · IQ2_XS Decode 113.15 · Prefill 569.16 ·
  37. Ollama · IQ3_S Decode 89.33 · Prefill 357.04 ·
  38. Ollama · Q2_K Decode 70.49 · Prefill 362.07 ·
  39. Strata · Q5_K_M Decode 70.01 · Prefill 293.99 ·

Fit check

Physics-based estimate from VRAM and memory bandwidth: pick a model to see which quantizations fit and the expected speed

VRAM is not recorded for this hardware; fit cannot be computed

Formula: tok/s = (bandwidth ÷ model size) × 0.55, adapted from llmfit; see run records for measured numbers

Key specs

Vendor
Apple
Memory bandwidth
68 GB/s

Run records

39 configurations

ModelFrameworkQuantizationDecodeGen timeSamples
Qwen3.5-4Bllama.cppIQ2_M88 tok/s—1
Qwen3.5-4Bllama.cppQ4_K_M55.6 tok/s—1
Qwen3.5-4BStrataQ4_K_S87.2 tok/s—1
Qwen3.5-9Bllama.cppQ3_K_M51.9 tok/s—1
Qwen3.5-9BStrataIQ1_S96.9 tok/s—1
Qwen3.5-0.8BvLLMQ4_K_S132.9 tok/s—1
Qwen3.5-0.8BStrataQ2_K231.4 tok/s—1
Qwen3.8-9BOllamaIQ3_XS50.4 tok/s—1
Qwen3.8-9BOllamaQ4_K_M39.1 tok/s—1
Qwen3-8BOllamaIQ2_XS64.6 tok/s—1
Qwen3-4Bllama.cppQ3_K_L65.6 tok/s—1
Qwen3-4Bllama.cppQ8_036.8 tok/s—1
Qwen3-4BOllamaQ2_K72.5 tok/s—1
Qwen3-1.7Bllama.cppIQ3_S97.7 tok/s—1
Qwen3-1.7Bllama.cppQ5_K_M75.8 tok/s—1
Qwen3-1.7BvLLMQ4_K_M84.3 tok/s—1
Qwen3-1.7BvLLMQ6_K72.5 tok/s—1
Qwen3-4B-Thinking-2507llama.cppIQ2_XS75.2 tok/s—1
Qwen3-4B-Thinking-2507llama.cppQ4_K_M59.4 tok/s—1
Qwen3-4B-Thinking-2507vLLMFP1630.1 tok/s—1
Qwen3-4B-Thinking-2507vLLMIQ2_XS79.9 tok/s—1
Qwen3-Coder-NextvLLMIQ4_XS47.9 tok/s—1
Qwen3-Coder-NextOllamaFP1621.4 tok/s—1
Qwen3-Coder-NextOllamaQ3_K_L41.5 tok/s—1
Qwen2.5-0.5B-Instructllama.cppIQ2_M189.8 tok/s—1
Qwen2.5-0.5B-Instructllama.cppQ2_K189.4 tok/s—1
Qwen2.5-0.5B-InstructvLLMQ3_K_S170.3 tok/s—1
Qwen2.5-0.5B-InstructvLLMQ4_K_M144 tok/s—1
Qwen2.5-Omni-3BOllamaIQ3_XS74.2 tok/s—1
Qwen2.5-Omni-3BOllamaQ6_K48.4 tok/s—1
Qwen2.5-Omni-3BStrataQ3_K_M114.6 tok/s—1
Qwen3-Embedding-0.6Bllama.cppIQ4_XS148.8 tok/s—1
Qwen3-Embedding-0.6BOllamaQ2_K166.6 tok/s—1
gemma-4-E2B-itllama.cppIQ2_XS125.4 tok/s—1
gemma-4-E2B-itllama.cppQ6_K62.5 tok/s—1
gemma-4-E2B-itOllamaIQ2_XS113.2 tok/s—1
gemma-4-E2B-itOllamaIQ3_S89.3 tok/s—1
gemma-4-E4B-itOllamaQ2_K70.5 tok/s—1
gemma-4-E4B-itStrataQ5_K_M70 tok/s—1