Zhipu API ARR hits $2B; inference costs drop 80% on domestic chips
xeophon · x · 2026-09-01
Zhipu's earnings report reveals massive growth: MaaS (API) ARR scaled from $250M in March to $2.0B (weekly runrate) by August. Token usage grew >40x YTD, and API pricing increased by 101%. API gross margin reached 24.6%, with inference costs per token dropping 80% and revenue per unit of compute increasing 14x. Zhipu is operating 100K domestic accelerators, with GLM-5.3 Flash processing 60-62T tokens in six days on domestic chips, achieving a 3x performance improvement via infra optimization. Beyond coding, the models are being trained for long-horizon planning and tool use, having already identified 2,436 vulnerabilities in cybersecurity. GLM-5.2 and 5.3 share the same base model, but 5.3 features heavier post-training.
More from Companies & People
- Tinyfool sparks debate over whether Tesla Roadster 2 counts as 'useless' innovation — tinyfool · 2026-10-16
- Tinyfool: does Xiaomi's foldable released days before iPhone count as innovation? — tinyfool · 2026-10-16
- US universities go on an AI philosopher hiring spree as philosophy departments add tenure-track AI roles — AnnaCiaunica · 2026-10-16
- NVIDIA and Mercedes-Benz Expand Partnership on Next-Gen Automated Driving — nvidia · 2026-10-16
- Ex-Amazon Engineer: AI Has Killed Bezos's Six-Page Memo Culture — bibryam · 2026-10-16
- DHH's Omarchy Shenzhen Meetup Draws 400+ With Demos on Xiaomi, ARM and iPhone Ports — LoganGrasby · 2026-10-16