On-demand deployments allow you to use Llama 2 7B on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits.

See the On-demand deployments guide for details.