benchmarking chinese models is hell
good luck waiting 1 hour for a single response and don't you dare sending 10 requests at a time
1:06 PM · Jun 1, 2026 · 8.2K Views
Lisan al Gaib, LisanBench creator, reports Chinese LLM APIs suffer from hour-long latencies and strict rate limits
Founder Bindu Reddy suggested self-hosting to bypass these limits.
--0--
Sentiment
Sentiment unavailable for this story.
Cluster Engagement
-
Views
-
Comments
-
Reposts
-
Bookmarks
Expand data
Posts from X
Most Activity
Most ActivityTimeline
VIEWS659LIKES7
Bindu Reddy@bindureddy
@scaling01 You have to host it 🔥
Lisan al Gaib@scaling01
benchmarking chinese models is hell
good luck waiting 1 hour for a single response and don't you dare sending 10 requests at a time
5hViews 659Likes 7Bookmarks 0