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
00.10.2
  1. 1
    Best Value: Radxa ROCK 2F 2GBRK3528A · $15
    0.19 tok/s/$
  2. 2
    0.16 tok/s/$
  3. 3
    Fastest Under $100: Radxa X4 8GBN100 · $100
    0.14 tok/s/$
  4. 4
    Dshan Pi A1 4GBRK3576 · $69
    0.13 tok/s/$
  5. 5
    0.13 tok/s/$
  6. 6
    Orange Pi 5B 4GBRK3588S · $75
    0.12 tok/s/$
  7. 7
    Raspberry Pi 5 2GBBCM2712 · $82
    0.12 tok/s/$
  8. 8
    0.11 tok/s/$
  9. 9
    Orange Pi 4A 4GBT527 · $50
    0.1 tok/s/$
  10. 10
    Radxa ROCK 2A 4GBRK3528A · $30
    0.1 tok/s/$
  11. 11
    Raspberry Pi 5 4GBBCM2712 · $118
    0.1 tok/s/$
  12. 12
    Radxa NX5 8GBRK3588S · $119
    0.09 tok/s/$
  13. 13
    Orange Pi 5 Pro 16GBRK3588S · $120
    0.08 tok/s/$
  14. 14
    Radxa CM5 Lite 8GBRK3582 · $119
    0.08 tok/s/$
  15. 15
    Radxa ROCK 5 ITX 8GBRK3588 · $120
    0.08 tok/s/$
  16. 16
    Radxa ROCK 5T 12GBRK3588 · $125
    0.08 tok/s/$
  17. 17
    Orange Pi 5 Plus 8GBRK3588 · $135
    0.07 tok/s/$
  18. 18
    0.07 tok/s/$
  19. 19
    Orange Pi CM4 2GBRK3566 · $40
    0.07 tok/s/$
  20. 20
    Radxa Dragon Q6A 6GBQCS6490 · $119
    0.07 tok/s/$
  21. 21
    Radxa ROCK 3C 2GBRK3566 · $40
    0.07 tok/s/$
  22. 22
    Radxa ROCK 4C+ 4GBRK3399-T · $80
    0.07 tok/s/$
  23. 23
    Radxa ROCK 4D 8GBRK3576 · $106
    0.07 tok/s/$
  24. 24
    Cool Pi 4B 4GBRK3588S · $140
    0.06 tok/s/$
  25. 25
    0.06 tok/s/$
  26. 26
    0.06 tok/s/$
  27. 27
    0.06 tok/s/$
  28. 28
    Radxa 4 SE 4GBRK3399-T · $80
    0.06 tok/s/$
  29. 29
    Raspberry Pi 400 4GBBCM2711 · $65
    0.06 tok/s/$
  30. 30
    Raspberry Pi 5 8GBBCM2712 · $189
    0.06 tok/s/$
  31. 31
    0.06 tok/s/$
  32. 32
    ArmSoM W3 8GBRK3588 · $316
    0.05 tok/s/$
  33. 33
    0.05 tok/s/$
  34. 34
    Raspberry Pi 4 2GBBCM2711 · $72
    0.05 tok/s/$
  35. 35
    Raspberry Pi 500 8GBBCM2712 · $194
    0.05 tok/s/$
  36. 36
    ArmSoM AIM7 8GBRK3588 · $297
    0.04 tok/s/$
  37. 37
    Banana Pi R4 8GBMT7988A · $146
    0.04 tok/s/$
  38. 38
    Forlinx FET3576-C 2GBRK3576 · $210
    0.04 tok/s/$
  39. 39
    LattePanda Mu 8GBN100 · $398
    0.04 tok/s/$
  40. 40
    0.04 tok/s/$
  41. 41
    Raspberry Pi 5 16GBBCM2712 · $329
    0.04 tok/s/$
  42. 42
    ArmSoM Sige5 8GBRK3576 · $239
    0.03 tok/s/$
  43. 43
    0.03 tok/s/$
  44. 44
    Luckfox Aura 4GBRV1126B · $101
    0.03 tok/s/$
  45. 45
    Orange Pi 3B 4GBRK3566 · $94
    0.03 tok/s/$
  46. 46
    Radxa Dragon Q8B 32GBSnapdragon 8cx Gen 3 · $677
    0.03 tok/s/$
  47. 47
    Fastest Overall: Radxa Fogwise Airbox Q900 36GBIQ-9075 · $849
    0.03 tok/s/$
  48. 48
    0.03 tok/s/$
  49. 49
    Radxa ROCK 4B+ 4GBRK3399 OP1 · $170
    0.03 tok/s/$
  50. 50
    Radxa Zero 3W 8GBRK3566 · $92
    0.03 tok/s/$
  51. 51
    0.02 tok/s/$
  52. 52
    Banana Pi M5 4GBS905X3 · $146
    0.02 tok/s/$
  53. 53
    DEBIX Model A 4GBiMX 8M Plus · $135
    0.02 tok/s/$
  54. 54
    DEBIX Model C 2GBiMX 93 · $65
    0.02 tok/s/$
  55. 55
    Radxa ROCK 5B+ 16GBRK3588 · $135
    0.02 tok/s/$
  56. 56
    Raspberry Pi 500+ 16GBBCM2712 · $443
    0.02 tok/s/$
  57. 57
    0.01 tok/s/$
  58. 58
    Banana Pi CM4 4GBA311D · $104
    0.01 tok/s/$
  59. 59
    0.01 tok/s/$
00.10.2
Value (tokens/s per dollar), higher is better
Showing 59 boards · median 0.06 tok/s/$ (7 without a Per Dollar result not shown)Top 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.