Comparisons of Tiny Open Source AI Models (<4B)

Open source AI models with 4 billion parameters or fewer. These are usually the smallest models in terms of resource demand. Models are considered Open Source (also commonly referred to as open weights) where their weights are accessible to download. This allows self-hosting on your own infrastructure and enables customizing the model such as through fine-tuning. Click on any model to see detailed metrics. For more details including relating to our methodology, see our FAQs.

Alibaba logoQwen3 1.7B (Reasoning) and Microsoft Azure logoPhi-4 Mini are the highest intelligence Tiny open source models, defined as those with <4B parameters, followed by Alibaba logoQwen3 1.7B & Alibaba logoQwen3 0.6B (Reasoning).

Highlights

Intelligence
Artificial Analysis Intelligence Index; Higher is better
Total Parameters
Trainable parameters in billions
Further details
WeightsProvider
Benchmarks
Alibaba logo
Qwen3 1.7B (Reasoning)
Alibaba
38
2.03B
32.0k
$0.4
138
🤗
Alibaba Cloud
View
Microsoft Azure logo
Phi-4 Mini Instruct
Microsoft Azure
26
3.84B
128k
-
28
🤗
Microsoft Azure
View
Alibaba logo
Qwen3 1.7B
Alibaba
25
2.03B
32.0k
$0.2
141
🤗
Alibaba Cloud
View
Alibaba logo
Qwen3 0.6B (Reasoning)
Alibaba
23
0.752B
32.0k
$0.4
229
🤗
Alibaba Cloud
View
Microsoft Azure logo
Phi-3 Mini Instruct 3.8B
Microsoft Azure
22
3.8B
4.10k
-
-
🤗
-
View
Meta logo
Llama 3.2 Instruct 3B
Meta
20
3B
128k
$0.0
124
🤗
Novita
Hyperbolic
Amazon Bedrock
+5 more
View
DeepSeek logo
DeepSeek R1 Distill Qwen 1.5B
DeepSeek
19
1.5B
128k
-
-
🤗
-
View
Alibaba logo
Qwen3 0.6B
Alibaba
17
0.752B
32.0k
$0.2
232
🤗
Alibaba Cloud
View
Google logo
Gemma 3 1B Instruct
Google
13
1B
32.0k
-
-
🤗
?
View
Meta logo
Llama 3.2 Instruct 1B
Meta
10
1B
128k
$0.1
181
🤗
Deepinfra
SambaNova
Amazon Bedrock
View