GLM-5.3 (open-weight) beat Anthropic/OpenAI models – for 1/5 the cost
What happened
Seventeen models on the Featherbench leaderboard: glm-5.3 leads at 100%, eight tie at 96%, and checker errors flip the safety ranking. See who tops the board.
Summary assembled by rule from the sources below
Why it's spreading
- DiscussionCommunity discussion only (1 posts), not yet corroborated
Timeline
- First appeared on Hacker NewsHacker News