Revolut deploys self-hosted GLM at tenth the cost of API providers
The fintech firm built its own inference infrastructure with Nvidia hardware, cutting token costs from ~$3 per million to 30 cents.
What to know
- Revolut reduced inference costs to roughly 30 cents per million tokens by self-hosting on 56 Nvidia B200s—a tenfold savings versus prior Anthropic and OpenAI API spending.
- The move required $2.5M–3M upfront hardware investment plus ongoing electricity and maintenance costs; breakeven depends on prior token spending levels.
- Engineers report productivity gains when token rationing constraints lifted, and open-source models can handle 90% of work with precise requirements engineering.
- Trade-offs include potential quality gaps on ambiguous prompts and codebase lock-in, balanced against data privacy and cost predictability advantages.
The dispute Whether open-model deployment signals a durable shift that will erode API provider valuations and harm junior developers, or represents a niche optimization that leaves proprietary models dominant for high-stakes work. · positions read across 21 posts and comments
Self-hosting and open models make economic and strategic sense for large companies seeking cost reduction and data sovereignty.
-
“We did the same thing, on much less powerful hardware but it has served us very well and we don't need to worry at all about the question whether or not the suppliers of inference services can be trusted long term with our data.”
jacquesm · Hacker News ↗
Proprietary models like Claude remain superior for quality-critical work despite higher costs, and self-hosting introduces operational and maintenance risks.
-
“If you were to take away my work's $500 budget for Opus 5.5 and replace it with unlimited use of GLM-5.3, it would be a huge downgrade in my opinion.”
user43928 · Hacker News ↗
Junior developers face genuine harm from this trend as companies prioritize cost over developer growth, with AI replacing entry-level work before wages adjust.
-
“"This technology will completely replace you and lead to your unemployment, so you should fund it out of your own pocket"? Is that what you're saying? If so, can you please explain your reasoning?”
bakugo · Hacker News ↗
David Abitbol AI ops lead, RevolutRevolut Fintech company
How it unfolded 1 development · click the chart to see its coverage articlespostscomments
-
1
Revolut announces self-hosted GLM infrastructure deployment
Revolut revealed it has moved from relying on Anthropic and OpenAI APIs to running its own large language model on self-hosted hardware. According to David Abitbol, who leads AI ops at the company, the infrastructure consists of 56 Nvidia B200s achieving a blended cost of 30 cents per million tokens—approximately a tenth of what Revolut previously paid to external API providers.
“Per David Abitbol, who leads AI ops there: 56 Nvidia B200s, a blended 30 cents per million tokens, roughly a tenth of what they paid Anthropic and OpenAI.”
— gosen, Hacker News commenter · source -
1 outlet Revolut moves to self-hosted GLM
first by HN Frontpage, 5d ago
-
"the gains showed up when tokens stopped being rationed" - I can totally see that. I am leading (as an external consultant) a larger dev team for a huge insurance company. The devs are quite junior or in general more the "I became a sw-eng here because that company is the only big employer in a 1h radius" type of people (meaning: No rockstar…
2 more of the top 3 · 21 posts in this stretch
-
> Per David Abitbol, who leads AI ops there: 56 Nvidia B200s, a blended 30 cents per million tokens, roughly a tenth of what they paid Anthropic and OpenAI.So they've invested around $2.5M-3M into hardware + monthly operating costs (electricity, updates, cooling). I wonder how much did they spend on tokens to decide to make such a switch. I mean…
-
Thanks for sharing!A year or eighteen months ago, if you were to ask me how much I would personally pay to have access to a then-frontier model for work, I would have said: not much. They were frustrating to instruct and use. The cost of play was too high.Six months ago, I think I would have paid $100 per month.Today, for a current frontier model…
-
What people are saying 16 voices from 1 site · best of 21 · verbatim
- At what scale and token spend does the $2.5M–3M hardware investment break even for a mid-sized company?
- Does GLM-5.3 maintain parity with Claude for regulatory-sensitive financial services work, or does Revolut accept a quality tradeoff?
- Will insurance and fintech companies increasingly demand in-house AI infrastructure for data sovereignty and compliance, creating a new operational burden?
- Nov 3
-
Would enjoy chatting offline if up for it. I am an external consultant acting as more of an operations/product lead to insurance, but come the insurance side.Curious about consulting model? Are you external (full-time) lead or fractional? Embedded in their tech and security stack or just helping organize/lead through communication channels?
-
I'm currently evaluating GLM and oh boy, lots of correctness issues on large contexts if you run it on vLLM (bugs in the engine itself). I wonder what their config is.
-
Hey Opus 9, you are an insurance carrier offering web-based policies and servuces across the EU. Available tools: german_notary_api, bafin_compliance.Here is €1B in initial capital. Go!
-
I would sacrifice a few benchmark points for higher tok/s actually, especially if more bandwidth as well so I can scale up swarms, then I can likely compensate for any drop and get more mileage out of it.
-
> most of their junior devwork can be replaced completely by AI - so they really should start to learn how to work with it. I recommended everybody to get a private 20€/month claude subscriptionI'm struggling to wrap my head around this. "This technology will completely replace you and lead to your unemployment, so you should fund it out of your…
-
Why doesn't your company just get each of them a 20€ subscription instead of paying per token?
-
>I recommended everybody to get a private 20€/month claude subscription and use it for toy projects.That's about 20€/month more than I care to give Anthropic. They already stole my data and now they want my money too? No thanks. I'll stick to running Gemma or Qwen on my Mac.
-
What does "blended $0.30 per 1M tokens" mean? Is that input or output or average across the two? (asking as am trying to benchmark that cost against public GLM providers on Openrouter)
-
> ... and IMHO most of them are in denial of the fact that their skills have mostly become worthless if they don't start to pivot to something valueable.But your comment says it several time: the valuable skill is being already a seasoned dev.Then AI is superpower.It's for the junior that things sucks big times.I see it all around me: senior devs…
-
Are you not worried about this being a sort of one-way street? Do you have any strategies to avoid ending up with a codebase you cannot work on without "AI"? when I think insurance company, I would assume the value seems to be more in the trust you build up over decades of with customers, not producing lots of code faster.
-
At dayjob we're also using self-hosted GLM-5.3 via our flavour of pi.dev but teams still have discretionary power over using Claude Code or not (there's a team's budget to spend on whatever the team decides). Some teams like mine still like to use Claude Code for a lot of the work but most stuff doesn't require the latest and greatest LLM and can…
-
If you were to take away my work's $500 budget for Opus 5.5 and replace it with unlimited use of GLM-5.3, it would be a huge downgrade in my opinion.If your employer is so cheap as to spend a $100 or less per month per engineer, while being large enough not to have access to the subscription plans, a move to self hosted or cheap providers does…
-
This is a sane version of what “sovereign” AI can look like for many. In Canada we’ve just had this national plan to overpay for an inferior AI product in the name of sovereignty. Even if GLM is slightly behind, this is likely close to the Pareto frontier.They also could consider putting this on openrouter when they have spare capacity to make the…
-
For the past several months about 90% of the work we do goes to open source models. If you ignore the noise around 'cheap' etc, you might find that with a little extra effort to more precisely define what you need not only do these models generally work as well, and they will generally more closely follow instructions - on longer timeframes this…
-
We did the same thing, on much less powerful hardware but it has served us very well and we don't need to worry at all about the question whether or not the suppliers of inference services can be trusted long term with our data. The answer to that question in my opinion is 'absolutely not'.
-
If investors knew how powerful open models are the AI companies valuation would tank overnight