Revolut Switches to Self-Hosted GLM for Coding Tasks

1 reports archived1 independent sourcesupdated 11/3/2026, 17:53:59

Synthesis & Latest Updates

1 Sources Cross-Validated

Fintech giant Revolut has migrated approximately 2,000 engineers from frontier APIs like Claude Code and Codex to a self-hosted GLM model for coding tasks. Leveraging 56 Nvidia B200 GPUs, the setup achieves a blended cost of $0.30 per million tokens—roughly one-tenth of previous API expenses—and reduced engineering lead time by 70%. This shift highlights how open-weight models are becoming the default when cost barriers vanish, posing a significant pricing challenge for proprietary AI labs.

LATEST/Fintech giant Revolut has migrated approximately 2,000 engineers from frontier APIs like Claude Code and Codex to a self-hosted GLM model for coding tasks. Leveraging 56 Nvidia B200 GPUs, the setup achieves a blended cost of $0.30 per million tokens—roughly one-tenth of previous API expenses—and reduced engineering lead time by 70%. This shift highlights how open-weight models are becoming the default when cost barriers vanish, posing a significant pricing challenge for proprietary AI labs.

Hourly heat curve

2 fully observed hours

Fewer than 3 fully observed hours; trend not plotted yet.

Coverage timeline

Total 1 reports · Latest first

Hacker News AIT2Nov 3, 17:53·78 pts
Revolut Switches to Self-Hosted GLM for Coding Tasks
- Revolut uses 56 Nvidia B200s to self-host GLM, cutting token costs to ~1/10th of Anthropic/OpenAI rates
- Removing token rationing led to ubiquitous AI usage among engineers, reducing engineering lead time by 70%
- Open-weight models are becoming the default while frontier APIs become exceptions, disrupting current pricing models
View Related Reports & Analysis Original Source