Locateanything 3B
NVIDIA's Locateanything 3B is a chat model that excels at vision-language grounding tasks, such as object localization and detection, with a generalist design supporting various domains. Its Parallel Box Decoding architecture enables efficient prediction of bounding box coordinates in a single step, preserving geometric consistency.
Input
Output
Context
33K
Max Output
8K
Parameters
3.8B
Technical Specifications
Model TypeChat
Context Window32,768 tokens
Max Output Tokens8,192 tokens
Parameters3.8B
Release DateMar 2, 2026
Training CutoffNot available
Licenseother
Open SourceYes
Input Modalities
TextImage
Output Modalities
Text
Capabilities
Resources & Links
Estimated GPU Requirements for Locateanything 3B
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Jun 23, 2026
Start building with the right model.
Automatically route workloads to the right model for every task, every time.