Deep dive into OllamaZone

Last refreshed Just now

Here's what's happening in our database. The stats are updated when new models are added or modified.

Browse All Models

Active Models

651

Ready to use right now

Model Types

6

Different families of models

Quantization Methods

8

Ways to run faster

Average Size of Models

5.3 GB

Not sure you really need this info

How We're Growing

Watch our server and models database expand day by day

Total New Models

651

Added over time period

Average Per Day

14.5

Models added daily

Peak Day

144 models

Apr 6

Popular Model Types

See what's trending in our collection

Llama

32.3%
211 models of 653 total
Llama
211
32.3% of total
Deepseek
168
25.7% of total
Other
84
12.9% of total
Gemma
76
11.6% of total
Qwen
75
11.5% of total
Mistral
39
6.0% of total

Model Family Mix

All LLM models, grouped by family, just for your information

llama

30.0%
195 models of 651 total
llama
195
30.0% of total
deepseek-r
140
21.5% of total
gemma
74
11.4% of total
qwen
51
7.8% of total
mistral
33
5.1% of total
Others
158
24.3% of total

Model Sizes

From compact to comprehensive

Parameter Sizes

44

Different model complexities

Most Common

1B

8 models

Avg Per Size

13

Models per parameter size

1B Parameter Models

Count: 8
Percentage: 1.4%

Global Reach

Where the models are running

Total Countries

19

Top Country

United States🇺🇸

% Top Country

33.8%

Hosting Providers & ISPs

Who's hosting the models

Total Providers

61

Top Provider

Hetzner Online GmbH

% Top Provider

15.7%

Quantization Methods

Making models faster and lighter

Total Methods

8

Most Used

Q4_K_M

% Most Used

59.9%

Storage Champions

The biggest models in OllamaZone collection

Largest Model

command-r-plus:latest