Snowflake CEO Finds GLM-5.2 Costs One-Fifth of Opus 4.7, Nears Performance
Decision Brief
What changedGLM-5.2 achieves near Claude Opus 4.7 performance on Snowflake benchmark at one-fifth the cost.
Why it mattersAI builders should note Chinese models challenging Western top models at drastically lower costs, potentially disrupting pricing and valuations.
Who should careTeams building on model APIs
Affected stackClaudeOpenAIGLM
Builder actionIgnore
Source confidenceMedium · Reliable media or first-hand reporting
Snowflake CEO tested a benchmark of 103 coding tasks and found that GLM-5.2 from Zhipu AI nearly matches Claude Opus 4.7, but costs only one-fifth per output token. However, GLM-5.2 consumes about twice as many tokens per task. Despite this, the enormous price gap pressures Anthropic and OpenAI, and could shake Western AI lab valuations.
Summary basis: official / RSS sourceUnless it says 'full article read', this summary is based only on publicly available content — it never pretends to have read restricted originals.
Sources
- The Decoder：AI News
- The Decoder：AI News