PaliGemma Models by Google DeepMind
Last refreshed 2026-05-19. Next refresh: weekly.
Details
Capabilities
About
PaliGemma is a family of open-source vision-language models (VLMs) developed by Google, emphasizing lightweight design and efficiency compared to other large language models. Built using open components, including the SigLIP vision model and the Gemma language model, PaliGemma models seamlessly process both images and text to deliver text outputs. This capability makes them well-suited for tasks such as image captioning, visual question answering, and object detection. Available in resolutions ranging from 224x224 to 896x896, these models are offered in various forms including pre-trained, mix, and fine-tuned versions to meet diverse research and practical needs. While useful for direct inference, they excel when fine-tuned for specific applications 13578.
Decision facts
- Best fit
- vision and multimodal work
- Capability starting point
- PaliGemma 3B 896 with 512 context and multimodal inputs
- Lowest tracked input
- Not tracked
- Closest related family
- T5Gemma
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
Use when the workload needs 512 context, 3B parameters, and multimodal inputs.
Use when the workload needs 512 context and 3B parameters.
Use when the workload needs 128 context and 3B parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| PaliGemma 3B 896 | Use when the workload needs 512 context, 3B parameters, and multimodal inputs. | 2024-05 | 512 context3B parametersmultimodal inputs | Current |
| PaliGemma 3B 448 | Use when the workload needs 512 context and 3B parameters. | 2024-05 | 512 context3B parameters | Current |
| PaliGemma 3B 224 | Use when the workload needs 128 context and 3B parameters. | 2024-05 | 128 context3B parameters | Current |
Release Timeline
1 release groupSpecifications(3 models)
| Model | Released | Context | Parameters | Vision | Multimodal |
|---|---|---|---|---|---|
| PaliGemma 3B 896 | 2024-05 | 512 | 3B | Yes | Yes |
| PaliGemma 3B 448 | 2024-05 | 512 | 3B | No | No |
| PaliGemma 3B 224 | 2024-05 | 128 | 3B | No | No |
