tools
Cerence CaLLM
Cerence CaLLM is a family of automotive language models for context-aware, multi-turn in-car conversations.
CaLLM provides cloud and embedded language models for automotive assistants. Its hybrid system handles multi-turn conversations, coordinates specialized AI agents, and supports vehicle functions, navigation, entertainment, and car information.
Automakers and Tier 1 suppliers use it to build branded in-car experiences. CaLLM Edge supports onboard processing when connectivity is limited. Cerence does not publish self-serve pricing; customers are directed to request a demo.
Features
- Cloud-based reasoning and agentic conversation management
- CaLLM Edge embedded processing for low-connectivity areas
- Coordinates responses across specialized AI agents
- Supports multi-turn automotive conversations
- Integrates OEM agents and components
- Supports vehicle controls, navigation, and point-of-interest search
- Includes automotive-specific guardrails and customization
Use cases
- Build branded voice assistants for vehicles
- Handle multi-step navigation and entertainment requests
- Process core vehicle interactions offline or with limited connectivity
- Answer vehicle-specific questions using OEM knowledge
- Coordinate specialized agents across an in-car conversation