> 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 221031

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

### On-Premises Release 221031: Docker Hub Images

- `deepgram/onprem-api:1.72.2`
- `deepgram/onprem-engine:3.37.8`
  - Minimum required NVIDIA driver version: `>=450.80.02`


- `deepgram/onprem-license-proxy:1.2.2`
- `deepgram/onprem-billing:1.4.0`
- `deepgram/onprem-metrics-server:2.0.0`

### Changes

- Deepgram On-premises users can now choose between Deepgram’s *Base* and *Enhanced* models in an ASR request via the `tier` query parameter, where `tier=base` will select the *Base *model and `tier=enhanced` will select the *Enhanced* model.
  - `tier` works in conjunction with the `detect_language` query parameter.
  - For users whose Enhanced models do not include the “*-enhanced” suffix in the model name, the use of the `tier` parameter is required.
  - Models may still be invoked directly via the model UUID without the use of `tier`.


- Deepgram On-premises deployments now support the following Understanding features (with the accompanying Understanding model deployed on-prem and the requisite configuration changes):
  - Topic Detection enables users to detect the most important and relevant topics that are referenced in speech within the audio.
`detect_topics=true&punctuate=true`
    - This requires the addition of the following section to the `api.toml` file:`[features]
  topic_detection = true`


  - When you use these Understanding features, please note that the punctuate=true parameter is required as part of the ASR request. If you do not explicitly include this parameter, it will be implicitly included by the system.


- Deepgram On-premises now supports the all-new “CloseStream” web socket message for closing your live audio streams. Please see the [New Methods for Closing Streams](https://deepgram.com/changelog/new-method-for-closing-streams/) changelog post for more information, or refer to the API documentation for [Transcribing Live Streaming Audio](https://developers.deepgram.com/api-reference/#transcription-streaming).