SenseTime's lightweight model reduces token consumption by 60%

robot
Abstract generation in progress

May 9, 2026

SenseTime Lightweight Model Token Consumption Reduced by 60%

=========================================================

SenseTime Technologies (00020) Releases Next-Generation Lightweight Multimodal Intelligent Model “SenseNova 6.7 Flash-Lite,” Featuring a Native Multimodal Architecture, Moving Away from the Traditional “Language + Vision” Concatenation Mode, and Token (Word Piece) Consumption is About 60% Lower Than Pure Text Intelligent Models, Effectively Reducing Commercial Computing Costs. SenseNova 6.7 Flash-Lite Uses Native Design …

(Excerpt)

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin