FitLocally by GPUwerk

Qwen · 30.5B · MoE

Can you run Qwen3-Coder-30B-A3B locally?

Qwen3-Coder-30B-A3B is a mixture-of-experts model with 30.5B parameters, 3.3B active per token. At 4-bit or better it runs on 73 of the 120 hardware configurations listed here; the smallest is the RTX 4090 (24 GB). On a 128 GB DGX Spark it runs at Q8_0, at roughly 43 tokens per second (our est.).

Data as of 2 November 2026

Fastest machines for Qwen3-Coder-30B-A3B

Every machine that runs it, fastest first. Filter by type; your machine is marked. Compare machines running it side by side

  1. 1RTX 5090Graphics card · 32 GB~287 tok/s ModelFit est.Q4_K_M · 20 of 31 GB
  2. 2RTX PRO 6000 BlackwellGraphics card · 96 GB~205 tok/s ModelFit est.Q8_0 · 34 of 95 GB
  3. 32x RTX 5090Graphics card · 64 GB~193 tok/s ModelFit est.Q8_0 · 34 of 62 GB
  4. 42x RTX PRO 6000 BlackwellGraphics card · 192 GB~193 tok/s ModelFit est.Q8_0 · 34 of 190 GB
  5. 5RTX 4090Graphics card · 24 GB~189 tok/s ModelFit est.Q4_K_M · 20 of 23 GB
  6. 6RTX 3090 TiGraphics card · 24 GB~175 tok/s ModelFit est.Q4_K_M · 20 of 23 GB
  7. 7RTX 3090Graphics card · 24 GB~166 tok/s ModelFit est.Q4_K_M · 20 of 23 GB
  8. 8RX 7900 XTXGraphics card · 24 GB~139 tok/s ModelFit est.Q4_K_M · 20 of 23 GB
  9. 9M5 UltraMac · 96–512 GB · 5 variants~130 tok/s ModelFit est.Q8_0 · 34 GB
  10. 102x RTX 4090Graphics card · 48 GB~122 tok/s ModelFit est.Q8_0 · 34 of 46 GB
  11. 112x RTX 3090 TiGraphics card · 48 GB~115 tok/s ModelFit est.Q8_0 · 34 of 46 GB
  12. 122x RTX 3090Graphics card · 48 GB~108 tok/s ModelFit est.Q8_0 · 34 of 46 GB
  13. 13M5 Max · 36 GB · 32cMac · 36 GB~97 tok/s ModelFit est.Q4_K_M · 20 of 24 GB
  14. 142x RX 7900 XTXGraphics card · 48 GB~97 tok/s ModelFit est.Q8_0 · 34 of 46 GB
  15. 15M3 UltraMac · 96–512 GB · 5 variants~90 tok/s ModelFit est.Q8_0 · 34 GB
  16. 16M4 Max · 36 GB · 32cMac · 36 GB~81 tok/s ModelFit est.Q4_K_M · 20 of 24 GB
  17. 17M2 UltraMac · 64–192 GB · 7 variants~80 tok/s ModelFit est.Q8_0 · 34 GB
  18. 18M5 MaxMac · 48–128 GB · 3 variants~80 tok/s ModelFit est.Q8_0 · 34 GB
  19. 19M2 MaxMac · 32 GB · 2 variants~76 tok/s ModelFit est.Q4_K_M · 20 GB
  20. 20M1 UltraMac · 64–128 GB · 4 variants~72 tok/s ModelFit est.Q8_0 · 34 GB
  21. 21M1 MaxMac · 32 GB · 2 variants~69 tok/s ModelFit est.Q4_K_M · 20 GB
  22. 22M4 MaxMac · 48–128 GB · 3 variants~67 tok/s ModelFit est.Q8_0 · 34 GB
  23. 23M3 Max · 36 GB · 30cMac · 36 GB~64 tok/s ModelFit est.Q4_K_M · 20 of 24 GB
  24. 24M3 MaxMac · 48–128 GB · 3 variants~52 tok/s ModelFit est.Q8_0 · 34 GB
  25. 25M2 MaxMac · 64–96 GB · 3 variants~49 tok/s ModelFit est.Q8_0 · 34 GB
  26. 26M5 ProMac · 48–64 GB · 4 variants~45 tok/s ModelFit est.Q8_0 · 34 GB
  27. 27M2 Pro · 32 GBMac · 32 GB~45 tok/s ModelFit est.Q4_K_M · 20 of 21 GB
  28. 28M1 MaxMac · 64 GB · 2 variants~44 tok/s ModelFit est.Q8_0 · 34 GB
  29. 29M6 · 32 GB · 12cMac · 32 GB~44 tok/s ModelFit est.Q4_K_M · 20 of 21 GB
  30. 30DGX Spark (GB10)AI mini PC · 128 GB~43 tok/s our est.Q8_0 · 34 of 122 GB
  31. 31M1 Pro · 32 GBMac · 32 GB~41 tok/s ModelFit est.Q4_K_M · 20 of 21 GB
  32. 32Ryzen AI Max+ 395 · 64 GBAI mini PC · 64 GB~40 tok/s our est.Q8_0 · 34 of 48 GB
  33. 33M3 Max · 96 GB · 30cMac · 96 GB~40 tok/s ModelFit est.Q8_0 · 34 of 72 GB
  34. 34M4 ProMac · 48–64 GB · 4 variants~38 tok/s ModelFit est.Q8_0 · 34 GB
  35. 35M5Mac · 32 GB · 2 variants~37 tok/s ModelFit est.Q4_K_M · 20 GB
  36. 36M3 Pro · 36 GBMac · 36 GB~36 tok/s ModelFit est.Q4_K_M · 20 of 24 GB
  37. 37Ryzen AI Max+ 395 · 128 GBAI mini PC · 128 GB~31 tok/s ModelFit est.Q8_0 · 34 of 96 GB
  38. 38M4 · 32 GBMac · 32 GB~28 tok/s ModelFit est.Q4_K_M · 20 of 21 GB
Too big for 47 machines

RX 7900 XT, M5 Pro · 24 GB · 20c, M5 Pro · 24 GB · 16c, M4 Pro · 24 GB · 20c, M4 Pro · 24 GB · 16c, M6 · 24 GB · 12c, M6 · 24 GB · 10c, M5 · 24 GB · 10c, M5 · 24 GB · 8c, M4 · 24 GB, M3 · 24 GB, M2 · 24 GB, RTX 5080, RTX 5070 Ti, RTX 5060 Ti, RTX 4080 SUPER, RTX 4080, RTX 4070 Ti SUPER, RTX 4060 Ti, Tesla T4, Tesla P100, RX 9070 XT, RX 9060 XT, M3 Pro · 18 GB, RTX 5070, RTX 4070 Ti, RTX 4070 SUPER, RTX 4070, RTX 3060, M2 Pro · 16 GB, M1 Pro · 16 GB, M6 · 16 GB · 12c, M6 · 16 GB · 10c, M5 · 16 GB · 10c, M5 · 16 GB · 8c, M4 · 16 GB, M3 · 16 GB, M2 · 16 GB, M1 · 16 GB, RTX 5060, RTX 4060, RTX 3070, RTX 3060 Ti, RTX 2070, M3 · 8 GB, M2 · 8 GB, M1 · 8 GB. Qwen3-Coder-30B-A3B needs about 20 GB at its smallest build.

Speeds are single-user decode estimates, not measurements. "ModelFit est." comes from ModelFit's model fitted on published runs (CC BY 4.0); "our est." is memory bandwidth divided by bytes read per token, times 0.55. Memory is what the best build needs, against what the machine can give a model. Method

DGX Spark setup notes: GPUwerk guide for Qwen3-Coder-30B-A3B.

About the model

Parameters
30.5B
Active per token
3.3B
Type
Mixture of experts
Good for
Coding assistant
Model card
Hugging Face
Hugging Face count
30.53B (agrees)

Builds and how to run them

QuantizationLoadMin. RAMCommand
Q4_K_M18.6 GB32 GBollama run qwen3-coder:30b-a3b-q4_K_M
Q8_032 GB48 GBollama run qwen3-coder:30b-a3b-q8_0

Load, minimum RAM and Ollama commands: ModelFit, CC BY 4.0. Load is memory with the model loaded, before long contexts.

Questions

Can I run Qwen3-Coder-30B-A3B on a Mac?

Yes on a Mac with enough memory: the smallest Apple configuration listed here that runs it at 4-bit or better is the M2 Max · 32 GB · 38c.

How much memory does Qwen3-Coder-30B-A3B need?

About 20 GB at the smallest build in our data, including a margin for the runtime and an 8K-token context. Longer contexts need more.

Does Qwen3-Coder-30B-A3B run on a DGX Spark?

Yes, at Q8_0. The Spark has 128 GB of unified memory, about 121.6 GB usable.