> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://developers.deepgram.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://developers.deepgram.com/_mcp/server.

## On-Premises Release 220708

Deepgram released a new version of its on-premises solution.

### On-Premises Release 220708 Docker Hub Images

- `deepgram/onprem-api:1.60.0`
- `deepgram/onprem-engine:3.28.10`
  - Minimum required NVIDIA driver version: `>=450.80.02`


- `deepgram/onprem-license-proxy:1.1.0`
- `deepgram/onprem-billing:1.4.0`
- `deepgram/onprem-metrics-server:2.0.0`

 

### This Release Contains The Following Changes

- Deepgram On-premises now supports Enhanced models for ASR requests.
  - Enhanced models can be invoked on-prem by using the `model=general-enhanced` request parameter. At this time, on-prem does not support the new `tier=enhanced` syntax.


- Transcription Find and Replace is now supported via the `replace` URI parameter. To learn more, please visit the feature page for [Find and Replace](https://developers.deepgram.com/documentation/features/replace/).
- The `/v2/models` endpoint response now includes the model UUID for cross-referencing against the model UUID listed in the `/v2/listen` response.

### Compatibility Notice

Any `[[post_processing_info.plugins]]` sections defined in the `api.toml` file will need to be deleted to get properly formatted responses with this release.