Perplexity unveils Portable Computer, a local-first agent

Portable Computer, a local-first agent with an on-device 27B model, scores 82.6% on real knowledge work, beating open-source harnesses Pi and Hermes; post-trained PPLX 27B reaches 85.4%. Parsing runs on-device, and escalation to a frontier advisor lifts Terminal Bench 2.1 score from 59.6% to 73.0% at $0.415 per rollout.