Europe's Biggest Tech Startup Cuts AI Coding Costs by 90% Using Chinese Model
Revolut, valued at $115 billion and Europe's most valuable private tech company, cut its AI coding costs to one-tenth of previous spending by self-hosting Z.ai's open-weight GLM model on dedicated Nvidia GPUs. The deployment reduced inference costs to 30 cents per million tokens versus commercial alternatives from Anthropic and OpenAI. The move also improved engineering lead times by 70% and coding time by 60% across 2,000 engineers. Revolut began testing commercial AI tools in January 2025 but shifted to self-hosting open-source models in April 2026 after concerns over escalating costs.
Key points
- Inference costs reduced to 30 cents per million tokens
- 90% reduction versus Anthropic and OpenAI spending
- 70% faster lead times and 60% faster coding time
- Self-hosted GLM model on Nvidia GPUs
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. Analytics India Magazine did the reporting; read the original for the full story.
Read the full story at Analytics India Magazine
