Models · The Decoder ·
Snowflake CEO finds GLM-5.2 competitive with Opus 4.7 at a fraction of the cost
Snowflake said Zhipu AI's GLM-5.2 performed close to Claude Opus 4.7 on a benchmark of 103 coding tasks at about one-fifth the cost per output token. The model used nearly twice as many tokens per task, but the price gap may increase competitive pressure on Anthropic and OpenAI.