From Dealroom · · 40s
Revolut shifts 2,000 engineers to self-hosted AI
Revolut cuts AI coding costs 90% by moving 2,000 engineers to self-hosted GLM models.
In brief
Revolut cuts AI coding costs 90% by moving 2,000 engineers to self-hosted GLM models. Revolut reduced AI coding costs by 90% after shifting 2,000 engineers from cloud AI to self-hosted GLM models. Originally reported by Dealroom.
Quick answers
- What's the deal?
- Revolut shifted roughly 2,000 engineers off Anthropic's Claude Code and OpenAI's Codex onto self-hosted GLM models, cutting AI coding costs by about 90%.
- Why now?
- Early rollout of Claude Code proved shaky economically, causing engineers to burn through large token budgets without clear returns.
- What is the key takeaway from Revolut shifts 2,000 engineers to self-hosted AI?
- Revolut reduced AI coding costs by 90% after shifting 2,000 engineers from cloud AI to self-hosted GLM models.
What's the deal?
Revolut shifted roughly 2,000 engineers off Anthropic's Claude Code and OpenAI's Codex onto self-hosted GLM models, cutting AI coding costs by about 90%.
Why now?
Early rollout of Claude Code proved shaky economically, causing engineers to burn through large token budgets without clear returns.
The endgame
Self-hosting GLM gives engineers freedom to use AI without unpredictable bills, improving coding speed and lead time.
At a glance
Revolut reduced AI coding costs by 90% after shifting 2,000 engineers from cloud AI to self-hosted GLM models.
