Gemma 3 4B is a compact AI model developed by Google DeepMind, offering an excellent balance between size and functionality. Unlike the base 1B version, the 4B model supports multimodal inputs: text, images (with resolution up to 896x896 pixels), and short videos. For example, the model can recognize objects and text in images, such as extracting data from receipts or labels. To process images, it uses the SigLIP visual encoder, which automatically segments large files.
Its innovative architecture and efficient 5:1 ratio of local-to-global attention optimize memory usage while supporting a context window of up to 128K tokens. Gemma 3 4B supports 35 languages, including Russian. The model also includes function calling capabilities, enabling integration with APIs for task automation, such as generating SQL queries or transforming data.
This model is ideal for creating intelligent assistants and for fast document and image processing, making it a perfect choice for developers who need multimodal capabilities without requiring significant computational resources.
| Model Name | Context | Type | GPU | Status | Link |
|---|
There are no public endpoints for this model yet.
An endpoint is a server with pre-installed software that automatically downloads neural network weights and launches an OpenAI-compatible API to handle incoming requests.
A private endpoint is created within the user's project, belongs exclusively to the user, and is billed per server configuration in accordance with the infrastructure pricing.
We recommend creating private endpoints when you need to:
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
131,072.0 |
1 | $0.38 | 25.45 | 1.504 | Launch | |
131,072.0 |
1 | $0.53 | 77.80 | 3.860 | Launch | |
131,072.0 |
1 | $0.83 | 119.26 | 4.176 | Launch | |
131,072.0 tensor |
2 | $0.97 | 111.86 | 2.496 | Launch | |
131,072.0 |
1 | $1.02 | 154.08 | 4.164 | Launch | |
131,072.0 tensor |
2 | $1.56 | 10.817 | Launch | ||
131,072.0 |
1 | $1.59 | 6.498 | Launch | ||
131,072.0 |
1 | $2.37 | 124.06 | 20.779 | Launch | |
131,072.0 |
1 | $3.83 | 141.00 | 14.260 | Launch | |
131,072.0 |
1 | $4.11 | 17.109 | Launch | ||
131,072.0 tensor |
2 | $4.61 | 44.024 | Launch | ||
131,072.0 |
1 | $4.74 | 26.675 | Launch | ||
131,072.0 tensor |
2 | $9.40 | 55.044 | Launch | ||
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
131,072.0 |
1 | $0.38 | 32.81 | 2.036 | Launch | |
131,072.0 |
1 | $0.53 | 86.02 | 4.362 | Launch | |
131,072.0 |
1 | $0.83 | 118.17 | 4.671 | Launch | |
131,072.0 tensor |
2 | $0.97 | 2.969 | Launch | ||
131,072.0 |
1 | $1.02 | 134.30 | 4.715 | Launch | |
131,072.0 tensor |
2 | $1.56 | 11.287 | Launch | ||
131,072.0 |
1 | $1.59 | 7.049 | Launch | ||
131,072.0 |
1 | $2.37 | 152.63 | 21.255 | Launch | |
131,072.0 |
1 | $3.83 | 191.74 | 14.044 | Launch | |
131,072.0 |
1 | $4.11 | 238.39 | 16.890 | Launch | |
131,072.0 tensor |
2 | $4.61 | 44.494 | Launch | ||
131,072.0 |
1 | $4.74 | 27.053 | Launch | ||
131,072.0 tensor |
2 | $9.40 | 55.367 | Launch | ||
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
131,072.0 |
1 | $0.38 | 17.61 | 1.002 | Launch | |
131,072.0 |
1 | $0.53 | 38.95 | 3.357 | Launch | |
131,072.0 |
1 | $0.83 | 53.83 | 3.842 | Launch | |
131,072.0 tensor |
2 | $0.97 | 1.809 | Launch | ||
131,072.0 |
1 | $1.02 | 68.01 | 3.661 | Launch | |
131,072.0 tensor |
2 | $1.56 | 10.127 | Launch | ||
131,072.0 |
1 | $1.59 | 88.28 | 5.996 | Launch | |
131,072.0 |
1 | $2.37 | 92.36 | 20.277 | Launch | |
131,072.0 |
1 | $3.83 | 107.33 | 13.274 | Launch | |
131,072.0 |
1 | $4.11 | 16.763 | Launch | ||
131,072.0 tensor |
2 | $4.61 | 43.334 | Launch | ||
131,072.0 |
1 | $4.74 | 26.330 | Launch | ||
131,072.0 tensor |
2 | $9.40 | 54.570 | Launch | ||
Contact our dedicated neural networks support team at support@immers.cloud or send your request to the sales department at sale@immers.cloud.