Posts tagged with "Azure GPU"

Revolut serves GLM-5.3 to about 2,000 engineers from 56 B200 GPUs at $0.30 per million tokens, a tenth of its previous Anthropic and OpenAI bill. On Azure the same model is available three ways: Claude in Foundry, GLM-5.3 per token via Fireworks, or the open weights on an H200 or MI300X node in Sweden Central, and only the last keeps prompts in Sweden, so the per-token and per-node break-even math below carries DORA and residency consequences as well as a price.