Open models, run in your account.
Exalt Reserve
Exalt Reserve runs open-weight models, self-hosted in your AWS, Azure, or Google Cloud account.
Five model families
- Llama
- Mistral
- Qwen
- Gemma
- DeepSeek
What self-hosted means here
In your account
The models run inside your cloud account, alongside retrieval, agents, and monitoring.
Chosen for the work
Which models run is decided with you for each use case. So is whether a use case also uses an outside AI service.
Open weights
Open-weight models, often called open source LLMs, have published weights, so they can run on infrastructure you choose.
Licenses
Each family is released under its publisher's own license terms.
The default is an open model, running in your own account.
Bring the work you want the models to do.
A first conversation about the work, the data it needs, and the cloud account it would run in.