GPU catalogue · Qualcomm

Qualcomm AI200

Rack-scale inference card and rack.

768 GB GPU memory Accelerator
Qualcomm AI200

Specifications

Qualcomm AI200: public facts

Good for: Serving large models at low cost per token.

MachineQualcomm AI200
Memory768 GB LPDDR per card
ArchitectureAI200
StatusAnnounced, expected 2026

Models

Which models fit the Qualcomm AI200

language models · image models · video models · embedding models.

Explore

More

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms