FitLocally by GPUwerk

Mistral · 7B

Can you run Mistral 7B Instruct locally?

Mistral 7B Instruct is a dense model with 7B parameters. At 4-bit or better it runs on 117 of the 120 hardware configurations listed here; the smallest is the RTX 5060 (8 GB). On a 128 GB DGX Spark it runs at Q8_0, at roughly 21 tokens per second (our est.).

Data as of 2 November 2026

Fastest machines for Mistral 7B Instruct

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~159 tok/s ModelFit est.Q8_0 · 8.4 of 31 GB
  2. 2RTX PRO 6000 BlackwellGraphics card · 96 GB~159 tok/s ModelFit est.Q8_0 · 8.4 of 95 GB
  3. 32x RTX 5090Graphics card · 64 GB~151 tok/s ModelFit est.Q8_0 · 8.4 of 62 GB
  4. 42x RTX PRO 6000 BlackwellGraphics card · 192 GB~151 tok/s ModelFit est.Q8_0 · 8.4 of 190 GB
  5. 5M5 UltraMac · 96–512 GB · 5 variants~100 tok/s ModelFit est.Q8_0 · 8.4 GB
  6. 6RTX 4090Graphics card · 24 GB~95 tok/s ModelFit est.Q8_0 · 8.4 of 23 GB
  7. 7RTX 5080Graphics card · 16 GB~94 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  8. 82x RTX 4090Graphics card · 48 GB~93 tok/s ModelFit est.Q8_0 · 8.4 of 46 GB
  9. 9RTX 3090 TiGraphics card · 24 GB~90 tok/s ModelFit est.Q8_0 · 8.4 of 23 GB
  10. 10RTX 5070 TiGraphics card · 16 GB~89 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  11. 112x RTX 3090 TiGraphics card · 48 GB~87 tok/s ModelFit est.Q8_0 · 8.4 of 46 GB
  12. 12RTX 3090Graphics card · 24 GB~84 tok/s ModelFit est.Q8_0 · 8.4 of 23 GB
  13. 132x RTX 3090Graphics card · 48 GB~82 tok/s ModelFit est.Q8_0 · 8.4 of 46 GB
  14. 14RX 7900 XTXGraphics card · 24 GB~77 tok/s ModelFit est.Q8_0 · 8.4 of 23 GB
  15. 152x RX 7900 XTXGraphics card · 48 GB~76 tok/s ModelFit est.Q8_0 · 8.4 of 46 GB
  16. 16RTX 4080 SUPERGraphics card · 16 GB~72 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  17. 17RTX 4080Graphics card · 16 GB~70 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  18. 18RTX 5070Graphics card · 12 GB~69 tok/s ModelFit est.Q8_0 · 8.4 of 11.0 GB
  19. 19M3 UltraMac · 96–512 GB · 5 variants~68 tok/s ModelFit est.Q8_0 · 8.4 GB
  20. 20RX 7900 XTGraphics card · 20 GB~67 tok/s ModelFit est.Q8_0 · 8.4 of 19.0 GB
  21. 21RTX 5060Graphics card · 8 GB~67 tok/s ModelFit est.Q5_K_M · 5.9 of 7.0 GB
  22. 22RTX 4070 Ti SUPERGraphics card · 16 GB~66 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  23. 23M2 UltraMac · 64–192 GB · 7 variants~61 tok/s ModelFit est.Q8_0 · 8.4 GB
  24. 24RTX 3070Graphics card · 8 GB~59 tok/s ModelFit est.Q5_K_M · 5.9 of 7.0 GB
  25. 25RTX 3060 TiGraphics card · 8 GB~59 tok/s ModelFit est.Q5_K_M · 5.9 of 7.0 GB
  26. 26M5 MaxMac · 48–128 GB · 3 variants~58 tok/s ModelFit est.Q8_0 · 8.4 GB
  27. 27RX 9070 XTGraphics card · 16 GB~58 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  28. 28M1 UltraMac · 64–128 GB · 4 variants~55 tok/s ModelFit est.Q8_0 · 8.4 GB
  29. 29RTX 2070Graphics card · 8 GB~53 tok/s ModelFit est.Q5_K_M · 5.9 of 7.0 GB
  30. 30RTX 4070 TiGraphics card · 12 GB~51 tok/s ModelFit est.Q8_0 · 8.4 of 11.0 GB
  31. 31RTX 4070 SUPERGraphics card · 12 GB~51 tok/s ModelFit est.Q8_0 · 8.4 of 11.0 GB
  32. 32RTX 4070Graphics card · 12 GB~51 tok/s ModelFit est.Q8_0 · 8.4 of 11.0 GB
  33. 33M4 MaxMac · 48–128 GB · 3 variants~49 tok/s ModelFit est.Q8_0 · 8.4 GB
  34. 34RTX 5060 TiGraphics card · 16 GB~47 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  35. 35M5 Max · 36 GB · 32cMac · 36 GB~44 tok/s ModelFit est.Q8_0 · 8.4 of 24 GB
  36. 36RTX 4060Graphics card · 8 GB~39 tok/s ModelFit est.Q5_K_M · 5.9 of 7.0 GB
  37. 37M3 MaxMac · 48–128 GB · 3 variants~38 tok/s ModelFit est.Q8_0 · 8.4 GB
  38. 38M4 Max · 36 GB · 32cMac · 36 GB~37 tok/s ModelFit est.Q8_0 · 8.4 of 24 GB
  39. 39M2 MaxMac · 32–96 GB · 5 variants~36 tok/s ModelFit est.Q8_0 · 8.4 GB
  40. 40RTX 3060Graphics card · 12 GB~34 tok/s ModelFit est.Q8_0 · 8.4 of 11.0 GB
  41. 41M5 ProMac · 24–64 GB · 6 variants~33 tok/s ModelFit est.Q8_0 · 8.4 GB
  42. 42M1 MaxMac · 32–64 GB · 4 variants~32 tok/s ModelFit est.Q8_0 · 8.4 GB
  43. 43RX 9060 XTGraphics card · 16 GB~32 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  44. 44RTX 4060 TiGraphics card · 16 GB~29 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  45. 45Tesla P100Graphics card · 16 GB~29 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  46. 46M3 MaxMac · 36–96 GB · 2 variants~29 tok/s ModelFit est.Q8_0 · 8.4 GB
  47. 47M4 ProMac · 24–64 GB · 6 variants~28 tok/s ModelFit est.Q8_0 · 8.4 GB
  48. 48Tesla T4Graphics card · 16 GB~25 tok/s ModelFit est.Q8_0 · 8.4 of 15.0 GB
  49. 49Ryzen AI Max+ 395 · 128 GBAI mini PC · 128 GB~23 tok/s ModelFit est.Q8_0 · 8.4 of 96 GB
  50. 50DGX Spark (GB10)AI mini PC · 128 GB~21 tok/s our est.Q8_0 · 8.4 of 122 GB
  51. 51M2 ProMac · 16–32 GB · 2 variants~20 tok/s ModelFit est.Q8_0 · 8.4 GB
  52. 52Ryzen AI Max+ 395 · 64 GBAI mini PC · 64 GB~20 tok/s our est.Q8_0 · 8.4 of 48 GB
  53. 53M6Mac · 24–32 GB · 3 variants~19 tok/s ModelFit est.Q8_0 · 8.4 GB
  54. 54M1 ProMac · 16–32 GB · 2 variants~18 tok/s ModelFit est.Q8_0 · 8.4 GB
  55. 55M6Mac · 16 GB · 2 variants~17 tok/s ModelFit est.Q8_0 · 8.4 GB
  56. 56M5Mac · 16–32 GB · 6 variants~16 tok/s ModelFit est.Q8_0 · 8.4 GB
  57. 57M3 ProMac · 18–36 GB · 2 variants~16 tok/s ModelFit est.Q8_0 · 8.4 GB
  58. 58M4Mac · 16–32 GB · 3 variants~12 tok/s ModelFit est.Q8_0 · 8.4 GB
  59. 59M3Mac · 16–24 GB · 2 variants~11 tok/s ModelFit est.Q8_0 · 8.4 GB
  60. 60M1 · 16 GBMac · 16 GB~10 tok/s ModelFit est.Q4_K_M · 6.6 of 10.7 GB
  61. 61M2Mac · 16–24 GB · 2 variants~10 tok/s ModelFit est.Q8_0 · 8.4 GB
Too big for 3 machines

M3 · 8 GB, M2 · 8 GB, M1 · 8 GB. Mistral 7B Instruct needs about 5.9 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

About the model

Parameters
7B
Active per token
7B
Type
Dense
Good for
Chat, Coding

Builds and how to run them

QuantizationLoadMin. RAMCommand
Q4_K_M5.5 GB8 GBollama run mistral:7b-instruct-v0.3-q4_K_M
Q5_K_M4.8 GB8 GBollama run mistral:7b-instruct-v0.3-q5_K_M
Q8_07.2 GB16 GBollama run mistral:7b-instruct-v0.3-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 Mistral 7B Instruct 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 Pro · 16 GB.

How much memory does Mistral 7B Instruct need?

About 5.9 GB at the smallest build in our data, including a margin for the runtime and an 8K-token context. Longer contexts need more.

Does Mistral 7B Instruct run on a DGX Spark?

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