Skip to content

Ollama LLM SBC Benchmark Rankings

Compare local large language model speed and efficiency across Single Board Computers

66 boards tested · speeds from 0.2 to 23.7 (Tokens/s, DeepSeek R1 1.5B)

66 results
Model

Full Results

Ranked by
01020
  1. 1
    Fastest Overall: Radxa Fogwise Airbox Q900 36GBIQ-9075 · $849
    23.7 tokens/s
  2. 2
    Radxa Dragon Q8B 32GBSnapdragon 8cx Gen 3 · $677
    21.9 tokens/s
  3. 3
    19 tokens/s
  4. 4
    16.9 tokens/s
  5. 5
    LattePanda Mu 8GBN100 · $398
    16.7 tokens/s
  6. 6
    Fastest Under $100: Radxa X4 8GBN100 · $100
    13.8 tokens/s
  7. 7
    Raspberry Pi 5 16GBBCM2712 · $329
    12 tokens/s
  8. 8
    Raspberry Pi 5 8GBBCM2712 · $189
    12 tokens/s
  9. 9
    Raspberry Pi 5 4GBBCM2712 · $118
    11.8 tokens/s
  10. 10
    Radxa NX5 8GBRK3588S · $119
    10.7 tokens/s
  11. 11
    Raspberry Pi 500+ 16GBBCM2712 · $443
    10.3 tokens/s
  12. 12
    Radxa CM5 Lite 8GBRK3582 · $119
    10.1 tokens/s
  13. 13
    Raspberry Pi 500 8GBBCM2712 · $194
    9.96 tokens/s
  14. 14
    Raspberry Pi 5 2GBBCM2712 · $82
    9.94 tokens/s
  15. 15
    Radxa ROCK 5T 12GBRK3588 · $125
    9.71 tokens/s
  16. 16
    ArmSoM AIM7 8GBRK3588 · $297
    9.65 tokens/s
  17. 17
    ArmSoM W3 8GBRK3588 · $316
    9.65 tokens/s
  18. 18
    9.6 tokens/s
  19. 19
    Orange Pi 5 Plus 8GBRK3588 · $135
    9.47 tokens/s
  20. 20
    Orange Pi 5B 4GBRK3588S · $75
    9.18 tokens/s
  21. 21
    Radxa ROCK 5 ITX 8GBRK3588 · $120
    9.15 tokens/s
  22. 22
    Orange Pi 5 Pro 16GBRK3588S · $120
    9.1 tokens/s
  23. 23
    Dshan Pi A1 4GBRK3576 · $69
    8.94 tokens/s
  24. 24
    Radxa Dragon Q6A 6GBQCS6490 · $119
    8.82 tokens/s
  25. 25
    Forlinx FET3576-C 2GBRK3576 · $210
    8.62 tokens/s
  26. 26
    ArmSoM Sige5 8GBRK3576 · $239
    8.22 tokens/s
  27. 27
    Cool Pi 4B 4GBRK3588S · $140
    8.15 tokens/s
  28. 28
    8.15 tokens/s
  29. 29
    7.78 tokens/s
  30. 30
    Radxa ROCK 4D 8GBRK3576 · $106
    7.69 tokens/s
  31. 31
    7.4 tokens/s
  32. 32
    Banana Pi R4 8GBMT7988A · $146
    5.73 tokens/s
  33. 33
    5.61 tokens/s
  34. 34
    Radxa ROCK 4C+ 4GBRK3399-T · $80
    5.27 tokens/s
  35. 35
    Radxa 4 SE 4GBRK3399-T · $80
    5.08 tokens/s
  36. 36
    Orange Pi 4A 4GBT527 · $50
    4.84 tokens/s
  37. 37
    Radxa ROCK 4B+ 4GBRK3399 OP1 · $170
    4.4 tokens/s
  38. 38
    Raspberry Pi 4 2GBBCM2711 · $72
    3.66 tokens/s
  39. 39
    Raspberry Pi 400 4GBBCM2711 · $65
    3.63 tokens/s
  40. 40
    3.61 tokens/s
  41. 41
    3.46 tokens/s
  42. 42
    Banana Pi M5 4GBS905X3 · $146
    3.01 tokens/s
  43. 43
    2.99 tokens/s
  44. 44
    Radxa ROCK 2A 4GBRK3528A · $30
    2.96 tokens/s
  45. 45
    Orange Pi 3B 4GBRK3566 · $94
    2.93 tokens/s
  46. 46
    DEBIX Model A 4GBiMX 8M Plus · $135
    2.85 tokens/s
  47. 47
    Best Value: Radxa ROCK 2F 2GBRK3528A · $15
    2.8 tokens/s
  48. 48
    Orange Pi CM4 2GBRK3566 · $40
    2.79 tokens/s
  49. 49
    Radxa ROCK 3C 2GBRK3566 · $40
    2.69 tokens/s
  50. 50
    Luckfox Aura 4GBRV1126B · $101
    2.67 tokens/s
  51. 51
    Radxa Zero 3W 8GBRK3566 · $92
    2.62 tokens/s
  52. 52
    2.44 tokens/s
  53. 53
    2.41 tokens/s
  54. 54
    2.35 tokens/s
  55. 55
    2.31 tokens/s
  56. 56
    2.19 tokens/s
  57. 57
    DEBIX Model C 2GBiMX 93 · $65
    1.4 tokens/s
  58. 58
    Banana Pi CM4 4GBA311D · $104
    1.35 tokens/s
  59. 59
    Radxa ROCK 5B+ 16GBRK3588 · $135
    1.12 tokens/s
  60. 60
    0.63 tokens/s
  61. 61
    0.42 tokens/s
  62. 62
    0.24 tokens/s
  63. 63
    Radxa ROCK 5B 8GBRK3588 · $160
    0.21 tokens/s
  64. 64
    ArmSoM Sige7 8GBRK3588 · $312
    0.2 tokens/s
  65. 65
    Radxa ROCK 5A 4GBRK3588S · $130
    0.2 tokens/s
  66. 66
    Radxa ROCK 5C 4GBRK3588S2 · $166
    0.2 tokens/s
01020
Generation Speed (tokens/s), higher is better
Showing 66 boards · median 5.61 tokens/sTop Picks from above

Frequently Asked Questions

What is the fastest single board computer in Ollama LLM (DeepSeek R1 1.5B)?

The Radxa Fogwise Airbox Q900 36GB currently tops our Ollama LLM (DeepSeek R1 1.5B) ranking with 23.7 (Tokens/s), measured on real hardware with our standardised test setup.

What is a good Ollama LLM (DeepSeek R1 1.5B) score for an SBC?

Across the 66 boards we have tested, Ollama LLM (DeepSeek R1 1.5B) results range from 0.2 to 23.7 (Tokens/s), with a median of 5.61. Anything above the median puts a board in the faster half of the market.

About Ollama LLM SBC Benchmark Rankings testing

This ranking shows how fast each board runs a large language model locally with Ollama. We use the DeepSeek R1 distilled models and run each one three times through the ollama-benchmark script, reporting the average generation speed (eval rate) in tokens per second: how quickly the answer appears, not counting prompt processing. Ollama runs on the CPU on these boards, so RAM size decides which models we can run: every board with around 2GB or more gets the 1.5B model, 8GB boards add the 8B model, and 16GB boards add 14B. Each model has its own ranking; switch between them above. Tokens per watt divides that speed by the board's peak power draw during the run, measured with a USB power meter, and tokens per dollar divides it by the board's current price. Higher is better for all three.