GITFIND

open-weight adoption — tracked daily

THE OPEN MODEL INDEX

Which open-weight models are being downloaded most? Hugging Face weight downloads and GitHub velocity for DeepSeek, Llama, Qwen, Kimi, Mistral, and GLM — plus the runtime layer that runs them. Every number says what it is; nothing here is made up.

30-day pulls: 8.9Mmodels tracked: 13 + 3 runtimesmost downloaded: glm 5.3 1.9M/30dfastest riser: kimi k3.1 pro · 1d

§ 1 — weight downloads, daily

fig. 01

5.0M10.0Mapr 30 — 4.7M commitsmay 2 — 154k commitsmay 7 — 795k commitsmay 8 — 4.7M commitsmay 9 — 463k commitsmay 10 — 580k commitsmay 11 — 795k commitsmay 13 — 693k commitsmay 15 — 396k commitsmay 16 — 947k commitsmay 19 — 4.9M commitsmay 21 — 5.3M commitsmay 23 — 592k commitsmay 25 — 4.4M commitsmay 27 — 4.7M commitsmay 29 — 138k commitsmay 30 — 5.0M commitsmay 31 — 1.0M commitsjun 1 — 682k commitsjun 2 — 328k commitsjun 4 — 7.4M commitsjun 6 — 5.7M commitsjun 7 — 5.7M commitsjun 8 — 5.7M commitsjun 11 — 6.0M commitsjun 15 — 218k commitsjun 16 — 100.0k commitsjun 17 — 558k commitsjun 18 — 204k commitsjun 19 — 996k commitsjun 21 — 6.8M commitsjun 24 — 7.0M commitsjun 29 — 696k commitsjun 30 — 171k commitsjul 1 — 8.8M commitsjul 3 — 8.0M commitsjul 4 — 151k commitsjul 5 — 745k commitsjul 7 — 1.1M commitsjul 9 — 8.6M commitsjul 10 — 11.1M commitsjul 11 — 175k commitsjul 12 — 8.7M commitsjul 14 — 142k commitsjul 15 — 10.8M commitsjul 16 — 0 commitsjul 17 — 237k commitsjul 18 — 930k commitsjul 19 — 82.9k commitsjul 20 — 8.6M commitsjul 21 — 16.4k commitsjul 22 — 10.1M commitsjul 23 — 16.4k commitsjul 24 — 208k commitsjul 26 — 130k commitsjul 28 — 195k commitsjul 29 — 11.4M commitsjul 30 — 11.4M commitsjul 31 — 11.9M commitsaug 1 — 12.2M commitsaug 2 — 12.0M commitsaug 3 — 12.5M commitsaug 4 — 12.6M commitsaug 5 — 13.1M commitsaug 6 — 13.1M commitsaug 7 — 13.0M commitsaug 8 — 12.8M commitsaug 9 — 12.7M commitsaug 10 — 12.3M commitsaug 11 — 12.2M commitsaug 12 — 12.2M commitsaug 13 — 12.1M commitsaug 14 — 12.4M commitsaug 15 — 12.2M commitsaug 16 — 12.0M commitsaug 17 — 11.8M commitsaug 18 — 11.6M commitsaug 19 — 11.5M commitsaug 20 — 11.3M commitsaug 21 — 10.8M commitsaug 22 — 10.3M commitsaug 23 — 10.1M commitsaug 24 — 9.8M commitsaug 25 — 9.7M commitsaug 26 — 9.6M commitsaug 27 — 8.5M commitsaug 28 — 7.7M commitsaug 29 — 7.7M commitsaug 30 — 7.3M commitsaug 31 — 7.2M commitssep 1 — 6.9M commitssep 2 — 6.9M commitssep 3 — 6.7M commitssep 4 — 6.0M commitssep 5 — 5.9M commitssep 6 — 6.3M commitssep 7 — 5.7M commitssep 8 — 5.7M commitssep 9 — 5.6M commitssep 10 — 5.6M commitssep 11 — 5.9M commitssep 12 — 6.3M commitssep 13 — 6.1M commitssep 14 — 5.8M commitssep 15 — 5.9M commitssep 16 — 6.5M commitssep 17 — 7.4M commitssep 18 — 7.1M commitssep 19 — 5.8M commitssep 20 — 6.2M commitssep 21 — 6.2M commitssep 22 — 6.6M commitssep 23 — 6.1M commitssep 24 — 7.6M commitssep 25 — 5.9M commitssep 26 — 6.4M commitssep 27 — 7.2M commitssep 28 — 7.8M commitssep 29 — 6.4M commitssep 30 — 6.4M commitsoct 1 — 2.1M commitsoct 2 — 750k commitsoct 4 — 834k commitsoct 5 — 1.7M commitsoct 6 — 3.2M commitsoct 9 — 2.8M commitsoct 10 — 1.3M commitsoct 11 — 3.1M commitsoct 12 — 3.1M commitsoct 13 — 1.3M commitsoct 15 — 1.3M commitsoct 16 — 1.2M commitsoct 17 — 1.3M commitsoct 18 — 1.9M commitsoct 19 — 3.2M commitsoct 20 — 1.4M commitsoct 21 — 3.3M commitsoct 23 — 3.4M commitsoct 24 — 5.0M commitsoct 25 — 1.6M commitsoct 26 — 3.7M commitsoct 27 — 3.9M commitsoct 28 — 4.1M commitsoct 29 — 2.4M commitsoct 30 — 2.4M commitsoct 31 — 4.1M commitsnov 1 — 26.1k commitsnov 2 — 4.6M commitsnov 3 — 4.6M commitsnov 4 — 1.9M commitsnov 5 — 797 commitsnov 6 — 3.6M commitsnov 7 — 5.5M commitsnov 8 — 5.5M commitsnov 9 — 9.2M commitsnov 10 — 9.1M commitsnov 11 — 9.1M commitsnov 12 — 8.9M commits8.9Mapr 30jun 2jul 7jul 30aug 18sep 6sep 25oct 18nov 12

hugging face trailing 30-day weight downloads per day, across all tracked models.

§ 2 — share of new downloads

fig. 02

22%20%18%18%22%
glm 5.3 21.8%deepseek v4.1 flash 19.8%inkling 18.5%deepseek r1 18.1%other (9) 21.8%

share of the trailing 30-day window. smaller models are grouped as “other”; each has its own chart below.

§ 3 — each model, same scale

fig. 03

glm 5.3

1.9M/30d · ▲ 1184%

deepseek v4.1 flash

1.8M/30d · ▲ 184%

inkling

1.6M/30d

deepseek r1

1.6M/30d · ▲ 104%

kimi k3

753k/30d · ▼ 69%

llama 3.3 70b

351k/30d · ▼ 58%

kimi k2

272k/30d · ▲ 69%

mistral small 3.1

158k/30d · ▼ 5%

deepseek v3

155k/30d · ▼ 85%

glm 4.5

116k/30d · ▲ 43%

qwen3 235b

79.6k/30d · ▼ 38%

kimi k3.1 pro

33.6k/30d · ▲ 4118%

qwen3.8

26.3k/30d · ▲ 45%

trailing 30-day weight downloads per day, averaged over 3-day buckets · all charts share one scale.

§ 4 — momentum, ranked

fig. 04

  • 01kimi k3.1 profastest

    33.6k/30d · ▲ 4118% · 1d to double · 0.4%

  • 02glm 5.3

    1.9M/30d · ▲ 1184% · 10d to double · 21.8%

  • 03deepseek v4.1 flash

    1.8M/30d · ▲ 184% · 28d to double · 19.8%

  • 04deepseek r1

    1.6M/30d · ▲ 104% · 89d to double · 18.1%

  • 05kimi k2

    272k/30d · ▲ 69% · 106d to double · 3.1%

  • 06qwen3.8

    26.3k/30d · ▲ 45% · 78d to double · 0.3%

  • 07glm 4.5

    116k/30d · ▲ 43% · 154d to double · 1.3%

  • 08inkling

    1.6M/30d · — · — · 18.5%

  • 09mistral small 3.1

    158k/30d · ▼ 5% · — · 1.8%

  • 10qwen3 235b

    79.6k/30d · ▼ 38% · — · 0.9%

  • 11llama 3.3 70b

    351k/30d · ▼ 58% · — · 3.9%

  • 12kimi k3

    753k/30d · ▼ 69% · — · 8.5%

  • 13deepseek v3

    155k/30d · ▼ 85% · — · 1.7%

§ 5 — the runtime layer

fig. 05

fig. 05 — the runtime layer (what people run open models with)

toolstarsforkscontributors
ollama187k19.0k629
llama.cpp134k25.5k2.1k
vllm96.8k24.7k2.4k

runtimes don’t publish weight downloads — repo traction is the honest proxy here.

§ 6 — intelligence brief

fig. 06

$ tail -f /var/log/gitfind/models.log

09:41:02 13 open models + 3 runtimes tracked — hf trailing-30d downloads + github velocity, nothing fabricated

09:41:03 glm 5.3 leads the 30-day window at 1.9M

09:41:04 glm 5.3 gained the most rate this month — +1.8M on the 30d window

09:41:05 kimi k3.1 pro is the fastest riser — doubling every 1 days at current pace

09:41:06 ollama leads the runtime layer at 187k★

09:41:07 router share coming when there’s a stable public source — we don’t scrape

09:41:08 █

form 27-b — subscription

THE TUESDAY BRIEFING

The repos that moved this week, why they matter, and what to watch next. One email. No noise.

building on this data? the catalog and the indexes are also an api — developers →