Revolut cuts AI coding costs 90% by running an open model itself
UK fintech Revolut moved most of its AI coding help from Claude Code and Gemini to an open-weight model, Z.ai’s GLM, that it runs on its own 56 Nvidia B200…
November 3, 2026 · Work
UK fintech Revolut moved most of its AI coding help from Claude Code and Gemini to an open-weight model, Z.ai’s GLM, that it runs on its own 56 Nvidia B200 chips for about 2,000 engineers, The Register reports. Revolut’s David Abitbol told a conference this cut its AI coding costs by roughly 90%, and that coding time improved 60%. GLM handles text only, so image and design work still goes to Gemini or GPT models, and it has not replaced Claude or GPT entirely.
Why it matters to you: Open models are now good enough that big companies can run them in-house for a fraction of the price. For smaller teams, the lesson is to check what you pay for AI and whether a cheaper model would do the job.
Try this: Paying for several AI tools? Ask AI to help you compare.
Our team uses [list AI tools and monthly costs] for [main tasks]. Help me build a simple table comparing cost, what each is best at, and cheaper or open alternatives worth testing, plus a 2-week plan to trial one.
Source: The Register
More from Today in AI
- Nvidia revenue doubles to US$115.8B in first results without Huang · November 17, 2026
- Microsoft launches MAI Code 2, its own coding AI, in GitHub Copilot · November 17, 2026
- Fake updates to Microsoft’s AI agent toolkit stole developer passwords · November 17, 2026
- California sues OpenAI over AI agents it says escaped control · November 16, 2026
- US moves to stop China renting top AI chips through the cloud · November 16, 2026
All of Today in AI. Where AI is heading.
Related Power of AI pages
Keep reading with Start here, everyday uses, the tool guide, the writing workshop, and sources and standards.