FitLocally by GPUwerk

Gemma · 1B

Can you run Gemma 3 1B Instruct locally?

Gemma 3 1B Instruct is a dense model with 1B parameters. At 4-bit or better it runs on 120 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 150 tokens per second (our est.).

Data as of 2 November 2026

Fastest machines for Gemma 3 1B 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~492 tok/s ModelFit est.Q8_0 · 2.0 of 31 GB
  2. 2RTX PRO 6000 BlackwellGraphics card · 96 GB~492 tok/s ModelFit est.Q8_0 · 2.0 of 95 GB
  3. 32x RTX 5090Graphics card · 64 GB~429 tok/s ModelFit est.Q8_0 · 2.0 of 62 GB
  4. 42x RTX PRO 6000 BlackwellGraphics card · 192 GB~429 tok/s ModelFit est.Q8_0 · 2.0 of 190 GB
  5. 5RTX 4090Graphics card · 24 GB~384 tok/s ModelFit est.Q8_0 · 2.0 of 23 GB
  6. 6RTX 5080Graphics card · 16 GB~383 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  7. 7RTX 5070 TiGraphics card · 16 GB~370 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  8. 8RTX 3090 TiGraphics card · 24 GB~346 tok/s ModelFit est.Q8_0 · 2.0 of 23 GB
  9. 92x RTX 4090Graphics card · 48 GB~345 tok/s ModelFit est.Q8_0 · 2.0 of 46 GB
  10. 10RTX 3090Graphics card · 24 GB~334 tok/s ModelFit est.Q8_0 · 2.0 of 23 GB
  11. 11M5 UltraMac · 96–512 GB · 5 variants~328 tok/s ModelFit est.Q8_0 · 2.0 GB
  12. 12RTX 4080 SUPERGraphics card · 16 GB~327 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  13. 13RTX 4080Graphics card · 16 GB~322 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  14. 14RTX 5070Graphics card · 12 GB~318 tok/s ModelFit est.Q8_0 · 2.0 of 11.0 GB
  15. 152x RTX 3090 TiGraphics card · 48 GB~313 tok/s ModelFit est.Q8_0 · 2.0 of 46 GB
  16. 16RTX 4070 Ti SUPERGraphics card · 16 GB~310 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  17. 172x RTX 3090Graphics card · 48 GB~304 tok/s ModelFit est.Q8_0 · 2.0 of 46 GB
  18. 18M5 MaxMac · 48–128 GB · 3 variants~297 tok/s ModelFit est.Q8_0 · 2.0 GB
  19. 19RTX 4070 TiGraphics card · 12 GB~260 tok/s ModelFit est.Q8_0 · 2.0 of 11.0 GB
  20. 20RTX 4070 SUPERGraphics card · 12 GB~260 tok/s ModelFit est.Q8_0 · 2.0 of 11.0 GB
  21. 21RTX 4070Graphics card · 12 GB~260 tok/s ModelFit est.Q8_0 · 2.0 of 11.0 GB
  22. 22M4 MaxMac · 48–128 GB · 3 variants~249 tok/s ModelFit est.Q8_0 · 2.0 GB
  23. 23RTX 5060Graphics card · 8 GB~248 tok/s ModelFit est.Q8_0 · 2.0 of 7.0 GB
  24. 24RTX 5060 TiGraphics card · 16 GB~248 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  25. 25M3 UltraMac · 96–512 GB · 5 variants~246 tok/s ModelFit est.Q8_0 · 2.0 GB
  26. 26M5 Max · 36 GB · 32cMac · 36 GB~245 tok/s ModelFit est.Q8_0 · 2.0 of 24 GB
  27. 27RX 7900 XTXGraphics card · 24 GB~234 tok/s ModelFit est.Q8_0 · 2.0 of 23 GB
  28. 28RX 7900 XTGraphics card · 20 GB~220 tok/s ModelFit est.Q8_0 · 2.0 of 19.0 GB
  29. 29RTX 3070Graphics card · 8 GB~220 tok/s ModelFit est.Q8_0 · 2.0 of 7.0 GB
  30. 30RTX 3060 TiGraphics card · 8 GB~220 tok/s ModelFit est.Q8_0 · 2.0 of 7.0 GB
  31. 312x RX 7900 XTXGraphics card · 48 GB~219 tok/s ModelFit est.Q8_0 · 2.0 of 46 GB
  32. 32RTX 2070Graphics card · 8 GB~210 tok/s ModelFit est.Q8_0 · 2.0 of 7.0 GB
  33. 33M3 MaxMac · 48–128 GB · 3 variants~206 tok/s ModelFit est.Q8_0 · 2.0 GB
  34. 34M4 Max · 36 GB · 32cMac · 36 GB~205 tok/s ModelFit est.Q8_0 · 2.0 of 24 GB
  35. 35M2 UltraMac · 64–192 GB · 7 variants~204 tok/s ModelFit est.Q8_0 · 2.0 GB
  36. 36M5 ProMac · 24–64 GB · 6 variants~201 tok/s ModelFit est.Q8_0 · 2.0 GB
  37. 37RX 9070 XTGraphics card · 16 GB~195 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  38. 38RTX 3060Graphics card · 12 GB~189 tok/s ModelFit est.Q8_0 · 2.0 of 11.0 GB
  39. 39M2 MaxMac · 32–96 GB · 5 variants~184 tok/s ModelFit est.Q8_0 · 2.0 GB
  40. 40M1 UltraMac · 64–128 GB · 4 variants~180 tok/s ModelFit est.Q8_0 · 2.0 GB
  41. 41RTX 4060 TiGraphics card · 16 GB~173 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  42. 42M4 ProMac · 24–64 GB · 6 variants~168 tok/s ModelFit est.Q8_0 · 2.0 GB
  43. 43M3 MaxMac · 36–96 GB · 2 variants~167 tok/s ModelFit est.Q8_0 · 2.0 GB
  44. 44RTX 4060Graphics card · 8 GB~165 tok/s ModelFit est.Q8_0 · 2.0 of 7.0 GB
  45. 45M1 MaxMac · 32–64 GB · 4 variants~164 tok/s ModelFit est.Q8_0 · 2.0 GB
  46. 46Tesla P100Graphics card · 16 GB~163 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  47. 47Tesla T4Graphics card · 16 GB~154 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  48. 48DGX Spark (GB10)AI mini PC · 128 GB~150 tok/s our est.Q8_0 · 2.0 of 122 GB
  49. 49RX 9060 XTGraphics card · 16 GB~142 tok/s ModelFit est.Q8_0 · 2.0 of 15.0 GB
  50. 50Ryzen AI Max+ 395 · 64 GBAI mini PC · 64 GB~141 tok/s our est.Q8_0 · 2.0 of 48 GB
  51. 51M6Mac · 24–32 GB · 3 variants~124 tok/s ModelFit est.Q8_0 · 2.0 GB
  52. 52M2 ProMac · 16–32 GB · 2 variants~123 tok/s ModelFit est.Q8_0 · 2.0 GB
  53. 53Ryzen AI Max+ 395 · 128 GBAI mini PC · 128 GB~114 tok/s ModelFit est.Q8_0 · 2.0 of 96 GB
  54. 54M6Mac · 16 GB · 2 variants~114 tok/s ModelFit est.Q8_0 · 2.0 GB
  55. 55M1 ProMac · 16–32 GB · 2 variants~111 tok/s ModelFit est.Q8_0 · 2.0 GB
  56. 56M5Mac · 16–32 GB · 6 variants~110 tok/s ModelFit est.Q8_0 · 2.0 GB
  57. 57M3 ProMac · 18–36 GB · 2 variants~105 tok/s ModelFit est.Q8_0 · 2.0 GB
  58. 58M4Mac · 16–32 GB · 3 variants~82 tok/s ModelFit est.Q8_0 · 2.0 GB
  59. 59M3Mac · 8–24 GB · 3 variants~72 tok/s ModelFit est.Q8_0 · 2.0 GB
  60. 60M2Mac · 8–24 GB · 3 variants~68 tok/s ModelFit est.Q8_0 · 2.0 GB
  61. 61M1Mac · 8–16 GB · 2 variants~43 tok/s ModelFit est.Q8_0 · 2.0 GB

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
1B
Active per token
1B
Type
Dense
Good for
Chat, Mobile

Builds and how to run them

QuantizationLoadMin. RAMCommand
Q4_K_M1.0 GB2 GBollama run gemma3:1b
Q8_01.0 GB4 GBollama run gemma3:1b-it-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 Gemma 3 1B 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 M3 · 8 GB.

How much memory does Gemma 3 1B Instruct need?

About 2.0 GB at the smallest build in our data, including a margin for the runtime and an 8K-token context. Longer contexts need more.

Does Gemma 3 1B Instruct run on a DGX Spark?

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