Compact AI models designed for local deployment of enterprise agents
Bolt is Nvidia's series of compact AI models built for local deployment of AI agents in enterprise environments. Released alongside the larger Nemotron 3.5 Super VL model, Bolt models are designed to run efficiently on-premises, enabling organizations to deploy AI agents that can operate continuously without relying on cloud infrastructure. These models complement Nvidia's broader enterprise AI strategy, offering a lighter-weight option for companies that need to run AI workloads locally while maintaining performance and reliability.
Who it's for
Enterprise AI developersOrganizations requiring on-premises AICompanies building AI agentsIT teams managing local AI infrastructure
Pricing · free
checked Oct 15, 2026
Plan
Price
Includes
Open Source
Free
Free to download and use · Designed for local deployment · Optimized for Nvidia hardware · Supports continuous AI agent operations
AI-researched pricing — verify on the official site before subscribing.
Use it for
— Deploying AI agents locally without cloud dependencies
Construct and deploy autonomous agents that run entirely on customer infrastructure with zero cloud dependencies
How the workflow runs
01muse-spark-open — Run lightweight reasoning tasks locally on Apple Silicon
02qwengyre — Train long-horizon planning agents using elastic RL framework
03bolt — Package agents for on-premises enterprise deployment
04spiral-coder — Generate deployment scripts and maintenance code locally
Rides the trend of enterprises demanding on-premises AI following security breaches at major providers. Entire stack runs without cloud services, addressing the growing market for air-gapped AI solutions.
