Is this page helpful?

Release Notes#

Release 2.0.0#

Highlights#

This is the first release of the newly-architected NIM VLM. This release ships with vLLM 0.19 as the sole inference backend.

This is the initial release of Nemotron-3-Content-Safety. For more information on this model, see the model card. For API examples, refer to the Vision Content Safety page.

Supported Hardware#

Refer to the support matrix for the following model:

Limitations#

  • Nemotron-3-Content-Safety

    • Image-only requests (no text prompt) are not supported and return HTTP 500. Always send a text prompt alongside the image.

    • Function (tool) calling, guided decoding, and reasoning are not supported.

For information about past updates and older versions, refer to the previous release notes.