Enable AI features
Enable Hyperglance's AI features by connecting up to an AI provider or using the local AI model
Hyperglance supports both an fully local AI model as well as external AI providers to power features such as:
-
Chat Assistant - Ask natural-language questions about your cost, inventory, and compliance data from any dashboard.
-
AI Search Generation - Turn a plain-language request into an advanced search query, or get an existing one explained.
-
AI Dashboard Generation - Describe the dashboard you want and let AI build it, complete with widgets.
-
AI Widget Configuration - Configure an individual widget by describing what you want it to show.
- AI agents - Task agents to build reports, monitor your environment, cleanup tags, send notifications and more. With intuitive & strict configurable guardrails to govern their behaviour.
Configure an AI provider
Go to Settings -> AI to configure your choice of provider with an API key. Hyperglance currently supports these providers:
- Anthropic (Claude)
- OpenAI
- Any OpenAI-compatible API endpoint
- Hyperglance's Local AI model (learn more below)

Hyperglance's Local AI model
The Local AI model option runs a small AI model within the Hyperglance deployment so that no data ever leaves your environment:
Note: We recommend hosting Hyperglance on a GPU-capable VM if you want to use the local model. Pure CPU inference does work but is very slow.
How to enable the local model
The local model is not available by default and must be enabled by your Hyperglance system administrator:
1) SSH into your Hyperglance deployment
2) Edit the /var/lib/data/hyperglance_helm_charts/values.yaml config file
3) Change ai.llama.enabled to true
ai:
llama:
enabled: true
4) Save the file and apply the changes using the hg CLI tool (this will restart Hyperglance)
hg apply-config
5. Once Hyperglance starts back up you can now select the "Hyperglance Local AI" option from the provider list on the Settings -> AI page and click Save: