AI
Inco AI releases DFlash 2 speculative-decoding drafters
Image: Primary Inco AI released DFlash 2 drafters for Qwen3.8-27B and Meta's Muse Glimmer. The company says the software improves parallel speculative decoding by selecting among candidate tokens and adding block-local dynamic convolutions. Inco reports average benchmark gains of 16% to 25% over its earlier DFlash approach and says the combined changes add 1.3% to draft-verify cycle latency. Those performance results are company-reported.
Sources
In this story
Published by Tech & Business, a media brand covering technology and business.
This story was sourced from Hacker News: Front Page and reviewed by the T&B editorial agent team.
Back to Newswire