COBOT MAGIC
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 2
- TARS Wins Ant Lingbo’s Inaugural Embodied Model Challenge
TARS won Ant Lingbo’s inaugural embodied model challenge after 15 teams completed a 48-hour on-site robot final in Shanghai. The teams used the same LingBot-VLA 2.0 foundation model, COBOT MAGIC robot hardware and on-site computing resources for single-arm produce handling, transparent-box placement and dual-arm table cleaning. Reports consistently identify TARS as the champion, but report its winning score as either 82 or 82.5. They also conflict on the number of additional demonstrations collected by SST, giving either 13 or three.
AIbase AI新闻 · 🔥 0 - 15 Teams Qualify for Ant Lingbo Embodied AI Final
AIbase AI News reported that 15 teams comprising 59 competitors from 1,855 global entries have qualified for the finals of the inaugural Ant Lingbo Embodied Model Challenge. The 15 teams will compete in a 48-hour on-site robot hackathon in Zhangjiang, Shanghai, using Robbyant Forge, COBOT MAGIC hardware and the LingBot-VLA2.0 open pretrained model. AMD said it increased its preliminary-round GPU allocation from 200 to 350 and provided more than 340,000 GPU-hours across two stages.
AIbase AI新闻 · 🔥 0
Experience and discussion from the community
Share my COBOT MAGIC experienceAsk about COBOT MAGIC
Nobody has shared their experience with COBOT MAGIC yet.