← LoLR HPCs
LoLR miniComing soon

Give materials, memory, and agents
a permanent local home

For single users and light workloads, a compact device hosts local models and personal knowledge bases, and connects with LoLR ticTALK series andLoLR.AI APP.

Computing architectureIntegrated SoCCurrent product definition; specific chip platform not yet disclosed
AI computing powerUp to 100 TOPS · INT8Theoretical peak; actual results depend on model, precision, and software stack
Memory64 GB high-bandwidth shared memoryShared memory, not 64 GB of dedicated GPU memory
Memory bandwidth150 GB/sCurrent product definition; subject to official datasheet and measurements
View specifications

Design philosophy

LoLR mini is designed to give individuals an always-available private AI environment with a smaller footprint and easier deployment, rather than replace heavy-duty workstations.

LoLR mini
Product images are for identification, not dimensional reference. Appearance is subject to official release and actual delivery.

Users & Scenarios

  • Individual professionalsTax, accounting, legal, and consulting document retrieval and drafting.
  • ManagersMeeting knowledge, project memory, and professional writing.
  • Researchers & CreatorsLiterature, notes, materials, and thematic synthesis.
  • Small officesAlways-on personal knowledge bases and lightweight agents.

Features & Value

Compact, always available

Built for continuous use on personal desks and in small offices.

Local space

Models, materials, conversations, and memory stay on the device.

Device collaboration

Locally process recordings from the ticTALK series and materials from LoLR.AI APP.

Lightweight tasks

Transcription, retrieval, summaries, and everyday content drafts.

Technical specifications

The table reflects the current product definition. Computing power is a theoretical peak; actual speed depends on the model and task.

ItemDescription
Computing architectureIntegrated SoCCurrent product definition; specific chip platform not yet disclosed
AI computing powerUp to 100 TOPS · INT8Theoretical peak; actual results depend on model, precision, context, and software stack
Memory64 GB high-bandwidth shared memoryShared memory, not 64 GB of dedicated GPU memory
Memory bandwidth150 GB/sCurrent product definition; subject to official datasheet and measurements
SoftwareLoLR.AI OS + Local models + Professional agentsCompatibility list, versions, and model benchmarks to be released
Chip platformNot yet disclosedStorage, power, networking, ports, weight, dimensions, and production appearance will be announced at release
About this page｜LoLR mini is coming soon. This page reflects the current product definition; final specifications and appearance are subject to official release.