GLM-5.3 (open-weight) beat Anthropic/OpenAI models – for 1/5 the cost
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
GLM-5.3 has outperformed other leading models from Anthropic and OpenAI while costing significantly less. The Featherbench leaderboard highlights its dominance, with several models closely trailing behind.
Comments (110 of ) ⌛