family member
Qwen2.5 logo

Qwen2.5 0.5B Instruct

248 tokens a second on Windows · GeForce RTX 3080 Ti 12 GB · 16 CPUs · 16 GB RAM. No hosted API sells it, so this network is the place to run it.

3 builds of Qwen2.5 0.5B Instruct, 18 signed runs on real machines: different packagers, precisions and engines, each a different file with its own numbers. 0.5B parameters, as the vendor states.

248tok/s, fastest machine
$0.56per M tokens · no API sells it
32Kcontext proved by recall
4.4 GBmemory, smallest build
9providers measured it
Run it on your gram Install a gram with one command. It works through the catalog on your own machine and signs every result. 9 providers have measured Qwen2.5 0.5B Instruct so far.Use it in your own room Ask Evo in your gramx and it runs on a machine that proved it, at the metered $0.56 per million tokens. On your own machines it costs you nothing.Compare every build Put it beside every other model in one table: speed, price, context and the API price, one row each.
best on ollama · 3 builds

trained 32K · proved 32K context, by Qwen Q4_K_M

247.9 tok/s peak, by Qwen Q4_K_M

Speed against price, every machine

Up and left is faster and cheaper · the line joins machines no other beats on both · price on a log scale

$1$10USD per million tokenstok/sWindows · GeForce RTX 3080 Ti 12 GB · 16 CPUs · 16 GB RAM · 4-bit by Qwen · 247.91 tok/s · $1.0352/MWindows · GeForce RTX 3080 Ti 12 GB · 16 CPUs · 16 GB RAM · 4-bit by Qwen · 247.91 tok/s · $1.0352/MLinux · NVIDIA 15 GB · 4 CPUs · 31 GB RAM · 4-bit by Qwen · 239.08 tok/s · $0.5645/MLinux · NVIDIA 15 GB · 4 CPUs · 31 GB RAM · 4-bit by Qwen · 239.08 tok/s · $0.5645/MWindows · Intel(R) UHD Graphics 730 2 GB · 8 CPUs · 16 GB RAM · 4-bit by Qwen · 27.68 tok/s · $2.578/MWindows · Intel(R) Iris(R) Xe Graphics 1 GB · 8 CPUs · 16 GB RAM · 4-bit by Qwen · 23.18 tok/s · $3.0787/MLinux · no GPU · 4 CPUs · 31 GB RAM · 4-bit by Qwen · 18.98 tok/s · $1.6229/MLinux · no GPU · 4 CPUs · 31 GB RAM · 4-bit by Qwen · 18.98 tok/s · $1.6229/MWindows · NVIDIA 4 GB · 12 CPUs · 7 GB RAM · 4-bit by Qwen · 5.75 tok/s · $37.5623/MWindows · NVIDIA 4 GB · 12 CPUs · 7 GB RAM · 4-bit by Qwen · 5.75 tok/s · $37.5623/M

Same build, different machines

Q4_K_M by Qwen on 9 machines: the fastest decodes 43× as fast as the slowest. One dot per machine, tokens per second.

  1. NVIDIA GeForce RTX 3080 Ti×1248 tok/s
  2. 2× Tesla T4×2239 tok/s
  3. Intel(R) UHD Graphics 730×132 tok/s
  4. Linux, CPU only · 5 CPUs×132 tok/s
  5. Apple silicon×131 tok/s
  6. Intel(R) Iris(R) Xe Graphics×131 tok/s
  7. Linux, CPU only · 6 CPUs×131 tok/s
  8. Windows, CPU only · 12 CPUs×15.8 tok/s

What it fits on

The smallest build needs about 4.4 GB; 15 of 18 kinds of machine on the network can hold it. Bars: each kind's memory (its card's, else RAM) · ticks: every build, at the size it held plus a fifth for the prompt.

  1. Linux, CPU only · 2 CPUs1 GB RAMnone
  2. NVIDIA GeForce RTX 3050 Laptop GPU4 GB on the cardnone
  3. NVIDIA GeForce GTX 16504 GB on the cardnone
  4. AMD Radeon(TM) Vega 8 Mobile Graphics7 GB RAMevery build
  5. Windows, CPU only · 12 CPUs7 GB RAMevery build
  6. Apple silicon8 GB RAMevery build
  7. Apple M18 GB RAMevery build
  8. Mac, CPU only · 8 CPUs8 GB RAMevery build
  9. NVIDIA GeForce RTX 3080 Ti12 GB on the cardevery build
  10. Tesla T415 GB on the cardevery build
  11. Intel(R) UHD Graphics 73016 GB RAMevery build
  12. Intel(R) Iris(R) Xe Graphics16 GB RAMevery build
  13. Apple M416 GB RAMevery build
  14. Apple M116 GB RAMevery build
  15. NVIDIA L422 GB on the cardevery build
  16. 2× Tesla T429 GB on the cardevery build
  17. Linux, CPU only · 4 CPUs31 GB RAMevery build
  18. NVIDIA RTX PRO 6000 Blackwell Server Edition95 GB on the cardevery build

How far back it remembers, machine by machine

Filled: the needle came back at that prompt size · outlined: asked and lost · blank: not asked · trained for 32K. 2 walks ended because the model forgot, 2 because the machine could not hold the prompt.

machine · build4K8K16K32K64K128Kwhy it stopped
Gram ·a5be462× Tesla T4 · Q4_K_M by Qwen32K recalled at every size asked
Gram ·a5be462× Tesla T4 · Q4_K_M by Qwen32K recalled at every size asked
Gram ·84bf54NVIDIA GeForce RTX 3080 Ti · Q4_K_M by Qwen8K the model lost the needle at 16K
Gram ·84bf54NVIDIA GeForce RTX 3080 Ti · Q4_K_M by Qwen8K the model lost the needle at 16K
Gram ·9e3c79Intel(R) UHD Graphics 730 · Q4_K_M by Qwen8K recalled at every size asked
Gram ·2e5c02Intel(R) Iris(R) Xe Graphics · Q4_K_M by Qwen4K recalled at every size asked
Gram ·e61d3b2× Tesla T4 · Q4_K_M by Qwen4K too slow to answer here at 8K
Gram ·e61d3b2× Tesla T4 · Q4_K_M by Qwen4K too slow to answer here at 8K
Gram ·9a1db0Windows, CPU only · 12 CPUs · Q4_K_M by Qwen— an older probe: where it stopped was not recorded
Gram ·9a1db0Windows, CPU only · 12 CPUs · Q4_K_M by Qwen— an older probe: where it stopped was not recorded
Gram ·d68672Apple silicon · by ollama— an older probe: where it stopped was not recorded

Carbon per million tokens

Lowest build on each machine · 2 of 5 read off a power sensor, the rest modelled from the parts · fewer grams is better

  1. Gram ·a5be4682 g0.77 J/tok · sensor
  2. Gram ·e61d3b98 g1.66 J/tok · modelled
  3. Gram ·9e3c79737 g3.75 J/tok · modelled
  4. Gram ·2e5c02936 g4.76 J/tok · modelled
  5. Gram ·84bf541.5 kg7.78 J/tok · sensor

Carbon is location-based: signed energy × the yearly average of the grid where the machine signed its place (United States grid average · EU grid average (applied to NL) · India grid average; Ember Global Electricity Review 2025 (2024 generation), location-based). Facility overhead is not included. A machine on its own solar is not credited until it can prove it.

Measured by

The people whose machines signed these numbers

proved by at least one build: Tool calling · Tool loop finished · Faithful to tool results · Structured output · Sustained output · Writes code · Follows instructions

Every build

Each its own file with its own page: the machines that ran it and their signatures.

buildenginewhose workmemorycontext provedbest tok/smachines
Q4_K_M by Qwenhf.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF:q4_k_mollamavendor · Qwen4.4 GB32K247.98
Q4_K_M by Qwenhf.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF:Q4_K_Mollamavendor · Qwen4.4 GB32K247.99
by ollamaqwen2.5:0.5bollamaollama——4.81