Dispatch
OpenAI Announces $200B Valuation Round   •   EU AI Act Compliance Deadline Extended to 2027   •   Google DeepMind Releases Gemini Ultra 3.0   •   Y Combinator S26 Batch: 60% of Startups Are AI-Native   •   MarTech Consolidation: Salesforce Acquires MadTech Pioneer   •   LLM Token Costs Drop 80% Year-Over-Year   •   Meta Llama 4 Released Under Permissive Commercial Licence   •   Anthropic's Claude Achieves New Benchmarks on Reasoning Tasks   •   Venture Capital Flows to AI Infrastructure Exceed $4B in Q2   •   Adobe GenStudio Reaches 500,000 Enterprise Users   •   OpenAI Announces $200B Valuation Round   •   EU AI Act Compliance Deadline Extended to 2027   •   Google DeepMind Releases Gemini Ultra 3.0   •   Y Combinator S26 Batch: 60% of Startups Are AI-Native   •   MarTech Consolidation: Salesforce Acquires MadTech Pioneer   •   LLM Token Costs Drop 80% Year-Over-Year   •   Meta Llama 4 Released Under Permissive Commercial Licence   •   Anthropic's Claude Achieves New Benchmarks on Reasoning Tasks   •   Venture Capital Flows to AI Infrastructure Exceed $4B in Q2   •   Adobe GenStudio Reaches 500,000 Enterprise Users
Est. MMXXV — Independent Digital PressWednesday, 2 September 2026Vol. I — No. 195
MarTech • Startups • LLMs • Digital Strategyterekhindigital.comMorning Edition

Terekhin Digital Media

Rigorous Journalism at the Frontier of Digital Commerce & Machine Intelligence

Wednesday, 2 September 2026Issue No. 195
LLMs

Perplexity and Nvidia Ship Portable Computer — On-Device AI Agents With No Per-Token Billing

The device runs AI agent workloads entirely locally, with zero token-cost billing for locally completed tasks and explicit user permission required before any escalation to cloud models — a direct challenge to cloud-hosted AI economics and data-residency constraints.

Perplexity and Nvidia have jointly released Portable Computer, a device designed to run AI agent workloads entirely on local hardware. Models, files, and processing operate on-device; tasks completed locally incur no token-cost billing. Escalation to cloud frontier models requires explicit user authorisation, establishing a privacy-by-default architecture rather than one that defaults to cloud transmission. The commercial proposition targets three enterprise segments with overlapping concerns: organisations with data-sovereignty requirements that preclude cloud transmission of sensitive content; high-inference-volume deployments where per-token costs at scale have become a material line item; and jurisdictions where cloud data-residency compliance introduces legal complexity. For each segment, a device-native architecture with no per-token billing represents a structurally different cost and risk model than cloud-first alternatives — and, notably, a distribution model that bypasses the API pricing dynamics of the major model providers entirely.

PerplexityNvidiaon-device AIportable computerinferencedata sovereignty
← Return to Front Page
Related Articles
© MMXXVI Terekhin Digital Media — All Rights Reserved — An Independent Digital Publication